Super	B-Malware
Mario	I-Malware
Run	I-Malware
Malware	I-Malware
#	O
2	O
–	O
DroidJack	B-Malware
RAT	I-Malware
Gamers	O
love	O
Mario	B-System
and	O
Pokemon	B-System
,	O
but	O
so	O
do	O
malware	O
authors	O
.	O

A	O
few	O
days	O
back	O
,	O
we	O
wrote	O
about	O
an	O
Android	B-System
Marcher	B-Malware
trojan	O
variant	O
posing	O
as	O
the	O
Super	B-System
Mario	I-System
Run	I-System
game	O
for	O
Android	B-System
.	O

We	O
have	O
found	O
another	O
instance	O
of	O
malware	O
posing	O
as	O
the	O
Super	B-System
Mario	I-System
Run	I-System
Android	B-System
app	O
,	O
and	O
this	O
time	O
it	O
has	O
taken	O
the	O
form	O
of	O
DroidJack	B-Malware
RAT	I-Malware
(	O
remote	O
access	O
trojan	O
)	O
.	O

Proofpoint	B-Organization
wrote	O
about	O
the	O
DroidJack	B-Malware
RAT	I-Malware
side-loaded	O
with	O
the	O
Pokemon	B-System
GO	I-System
app	O
back	O
in	O
July	O
2016	O
;	O
the	O
difference	O
here	O
is	O
that	O
there	O
is	O
no	O
game	O
included	O
in	O
the	O
malicious	O
package	O
.	O

The	O
authors	O
are	O
trying	O
to	O
latch	O
onto	O
the	O
popularity	O
of	O
the	O
Super	B-System
Mario	I-System
Run	I-System
game	O
to	O
target	O
eagerly	O
waiting	O
Android	B-System
users	O
.	O

Details	O
:	O
Name	O
:	O
Super	B-System
Mario	I-System
Run	I-System
Package	O
Name	O
:	O
net.droidjack.server	B-Indicator
MD5	O
:	O
69b4b32e4636f1981841cbbe3b927560	B-Indicator
Technical	O
Analysis	O
:	O
The	O
malicious	O
package	O
claims	O
to	O
be	O
the	O
Super	B-System
Mario	I-System
Run	I-System
game	O
,	O
as	O
shown	O
in	O
the	O
permissions	O
screenshot	O
below	O
,	O
but	O
in	O
reality	O
this	O
is	O
a	O
malicious	O
RAT	O
called	O
DroidJack	B-Malware
(	O
also	O
known	O
as	O
SandroRAT	B-Malware
)	O
that	O
is	O
getting	O
installed	O
.	O

Once	O
installed	O
,	O
the	O
RAT	O
registers	O
the	O
infected	O
device	O
as	O
shown	O
below	O
.	O

DroidJack	B-Malware
RAT	I-Malware
starts	O
capturing	O
sensitive	O
information	O
like	O
call	O
data	O
,	O
SMS	O
data	O
,	O
videos	O
,	O
photos	O
,	O
etc	O
.	O

Observe	O
below	O
the	O
code	O
routine	O
for	O
call	O
recording	O
.	O

This	O
RAT	O
records	O
all	O
the	O
calls	O
and	O
stores	O
the	O
recording	O
to	O
an	O
“	O
.amr	B-Indicator
”	O
file	O
.	O

The	O
following	O
is	O
the	O
code	O
routine	O
for	O
video	O
capturing	O
.	O

Here	O
,	O
the	O
RAT	O
stores	O
all	O
the	O
captured	O
videos	O
in	O
a	O
“	O
video.3gp	B-Indicator
”	O
file	O
.	O

It	O
also	O
harvests	O
call	O
details	O
and	O
SMS	O
logs	O
as	O
shown	O
below	O
.	O

Upon	O
further	O
inspection	O
,	O
we	O
have	O
observed	O
that	O
this	O
RAT	O
extracts	O
WhatsApp	B-System
data	O
too	O
.	O

The	O
RAT	O
stores	O
all	O
the	O
data	O
in	O
a	O
database	O
(	O
DB	O
)	O
in	O
order	O
to	O
send	O
it	O
to	O
the	O
Command	O
&	O
Control	O
(	O
C	O
&	O
C	O
)	O
server	O
.	O

The	O
following	O
are	O
the	O
DBs	O
created	O
and	O
maintained	O
by	O
the	O
RAT	O
.	O

We	O
saw	O
the	O
following	O
hardcoded	O
C	O
&	O
C	O
server	O
location	O
in	O
the	O
RAT	O
package	O
:	O
Conclusion	O
:	O
The	O
DroidJack	B-Malware
RAT	I-Malware
is	O
another	O
example	O
of	O
a	O
growing	O
trend	O
in	O
which	O
malware	O
authors	O
seek	O
to	O
exploit	O
public	O
interest	O
as	O
a	O
way	O
to	O
spread	O
malware	O
.	O

In	O
this	O
case	O
,	O
like	O
others	O
before	O
,	O
the	O
event	O
of	O
a	O
popular	O
game	O
release	O
became	O
an	O
opportunity	O
to	O
trick	O
unsuspecting	O
users	O
into	O
downloading	O
the	O
RAT	O
.	O

As	O
a	O
reminder	O
,	O
it	O
is	O
always	O
a	O
good	O
practice	O
to	O
download	O
apps	O
only	O
from	O
trusted	O
app	O
stores	O
such	O
as	O
Google	B-System
Play	I-System
.	O

This	O
practice	O
can	O
be	O
enforced	O
by	O
unchecking	O
the	O
"	O
Unknown	O
Sources	O
''	O
option	O
under	O
the	O
"	O
Security	O
''	O
settings	O
of	O
your	O
device	O
.	O

XLoader	B-Malware
Disguises	O
as	O
Android	B-System
Apps	O
,	O
Has	O
FakeSpy	B-Malware
Links	O
This	O
new	O
XLoader	B-Malware
variant	O
poses	O
as	O
a	O
security	O
app	O
for	O
Android	B-System
devices	O
,	O
and	O
uses	O
a	O
malicious	O
iOS	B-System
profile	O
to	O
affect	O
iPhone	B-System
and	O
iPad	B-System
devices	O
.	O

By	O
:	O
Hara	O
Hiroaki	O
,	O
Lilang	O
Wu	O
,	O
Lorin	O
Wu	O
April	O
02	O
,	O
2019	O
In	O
previous	O
attacks	O
,	O
XLoader	B-Malware
posed	O
as	O
Facebook	B-System
,	O
Chrome	B-System
and	O
other	O
legitimate	O
applications	O
to	O
trick	O
users	O
into	O
downloading	O
its	O
malicious	O
app	O
.	O

Trend	B-Organization
Micro	I-Organization
researchers	O
found	O
a	O
new	O
variant	O
that	O
uses	O
a	O
different	O
way	O
to	O
lure	O
users	O
.	O

This	O
new	O
XLoader	B-Malware
variant	O
poses	O
as	O
a	O
security	O
app	O
for	O
Android	B-System
devices	O
,	O
and	O
uses	O
a	O
malicious	O
iOS	B-System
profile	O
to	O
affect	O
iPhone	B-System
and	O
iPad	B-System
devices	O
.	O

Aside	O
from	O
a	O
change	O
in	O
its	O
deployment	O
techniques	O
,	O
a	O
few	O
changes	O
in	O
its	O
code	O
set	O
it	O
apart	O
from	O
its	O
previous	O
versions	O
.	O

This	O
newest	O
variant	O
has	O
been	O
labeled	O
XLoader	B-Malware
version	O
6.0	O
(	O
detected	O
as	O
AndroidOS_XLoader.HRXD	B-Indicator
)	O
,	O
following	O
the	O
last	O
version	O
discussed	O
in	O
a	O
previous	O
research	O
on	O
the	O
malware	O
family	O
.	O

Infection	O
chain	O
The	O
threat	O
actors	O
behind	O
this	O
version	O
used	O
several	O
fake	O
websites	O
as	O
their	O
host	O
—	O
copying	O
that	O
of	O
a	O
Japanese	O
mobile	O
phone	O
operator	O
’	O
s	O
website	O
in	O
particular	O
—	O
to	O
trick	O
users	O
into	O
downloading	O
the	O
fake	O
security	O
Android	B-System
application	O
package	O
(	O
APK	O
)	O
.	O

Monitoring	O
efforts	O
on	O
this	O
new	O
variant	O
revealed	O
that	O
the	O
malicious	O
websites	O
are	O
spread	O
through	O
smishing	O
.	O

The	O
infection	O
has	O
not	O
spread	O
very	O
widely	O
at	O
the	O
time	O
of	O
writing	O
,	O
but	O
we	O
’	O
ve	O
seen	O
that	O
many	O
users	O
have	O
already	O
received	O
its	O
SMS	O
content	O
.	O

In	O
the	O
past	O
,	O
XLoader	B-Malware
showed	O
the	O
ability	O
to	O
mine	O
cryptocurrency	O
on	O
PCs	O
and	O
perform	O
account	O
phishing	O
on	O
iOS	B-System
devices	O
.	O

This	O
new	O
wave	O
also	O
presents	O
unique	O
attack	O
vectors	O
based	O
on	O
the	O
kind	O
of	O
device	O
it	O
has	O
accessed	O
.	O

In	O
the	O
case	O
of	O
Android	B-System
devices	O
,	O
accessing	O
the	O
malicious	O
website	O
or	O
pressing	O
any	O
of	O
the	O
buttons	O
will	O
prompt	O
the	O
download	O
of	O
the	O
APK	O
.	O

However	O
,	O
successfully	O
installing	O
this	O
malicious	O
APK	O
requires	O
that	O
the	O
user	O
has	O
allowed	O
the	O
installation	O
of	O
such	O
apps	O
as	O
controlled	O
in	O
the	O
Unknown	O
Sources	O
settings	O
.	O

If	O
users	O
allow	O
such	O
apps	O
to	O
be	O
installed	O
,	O
then	O
it	O
can	O
be	O
actively	O
installed	O
on	O
the	O
victim	O
’	O
s	O
device	O
.	O

The	O
infection	O
chain	O
is	O
slightly	O
more	O
roundabout	O
in	O
the	O
case	O
of	O
Apple	B-System
devices	O
.	O

Accessing	O
the	O
same	O
malicious	O
site	O
would	O
redirect	O
its	O
user	O
to	O
another	O
malicious	O
website	O
(	O
hxxp	B-Indicator
:	I-Indicator
//apple-icloud	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
qwq-japan	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
or	O
hxxp	B-Indicator
:	I-Indicator
//apple-icloud	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
zqo-japan	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
)	O
that	O
prompts	O
the	O
user	O
to	O
install	O
a	O
malicious	O
iOS	B-System
configuration	O
profile	O
to	O
solve	O
a	O
network	O
issue	O
preventing	O
the	O
site	O
to	O
load	O
.	O

If	O
the	O
user	O
installs	O
the	O
profile	O
,	O
the	O
malicious	O
website	O
will	O
open	O
,	O
revealing	O
it	O
to	O
be	O
an	O
Apple	B-Organization
phishing	O
site	O
,	O
as	O
seen	O
in	O
figure	O
2	O
.	O

Technical	O
analysis	O
Most	O
of	O
this	O
new	O
attack	O
’	O
s	O
routines	O
are	O
similar	O
to	O
those	O
of	O
the	O
previous	O
XLoader	B-Malware
versions	O
.	O

However	O
,	O
as	O
mentioned	O
earlier	O
,	O
an	O
analysis	O
of	O
this	O
new	O
variant	O
showed	O
some	O
changes	O
in	O
its	O
code	O
in	O
line	O
with	O
its	O
new	O
deployment	O
method	O
.	O

We	O
discuss	O
these	O
changes	O
and	O
its	O
effect	O
on	O
Android	B-System
and	O
Apple	B-System
devices	O
.	O

Malicious	O
APK	O
Like	O
its	O
previous	O
versions	O
,	O
XLoader	B-Malware
6.0	I-Malware
abuses	O
social	O
media	O
user	O
profiles	O
to	O
hide	O
its	O
real	O
C	O
&	O
C	O
addresses	O
,	O
but	O
this	O
time	O
its	O
threat	O
actors	O
chose	O
the	O
social	O
media	O
platform	O
Twitter	B-Organization
,	O
which	O
was	O
never	O
used	O
in	O
previous	O
attacks	O
.	O

The	O
real	O
C	O
&	O
C	O
address	O
is	O
encoded	O
in	O
the	O
Twitter	B-Organization
names	O
,	O
and	O
can	O
only	O
be	O
revealed	O
once	O
decoded	O
.	O

This	O
adds	O
an	O
extra	O
layer	O
against	O
detection	O
.	O

The	O
code	O
for	O
this	O
characteristic	O
and	O
the	O
corresponding	O
Twitter	B-Organization
accounts	O
can	O
be	O
seen	O
in	O
figures	O
3	O
and	O
4	O
respectively	O
.	O

Version	O
6.0	O
also	O
adds	O
a	O
command	O
called	O
“	O
getPhoneState	O
”	O
,	O
which	O
collects	O
unique	O
identifiers	O
of	O
mobile	O
devices	O
such	O
as	O
IMSI	O
,	O
ICCID	O
,	O
Android	B-System
ID	O
,	O
and	O
device	O
serial	O
number	O
.	O

This	O
addition	O
is	O
seen	O
in	O
Figure	O
5	O
.	O

Considering	O
the	O
other	O
malicious	O
behaviors	O
of	O
XLoader	B-Malware
,	O
this	O
added	O
operation	O
could	O
be	O
very	O
dangerous	O
as	O
threat	O
actors	O
can	O
use	O
it	O
to	O
perform	O
targeted	O
attacks	O
.	O

Malicious	O
iOS	B-System
profile	O
In	O
the	O
case	O
of	O
Apple	B-System
devices	O
,	O
the	O
downloaded	O
malicious	O
iOS	B-System
profile	O
gathers	O
the	O
following	O
:	O
Unique	O
device	O
identifier	O
(	O
UDID	O
)	O
International	O
Mobile	O
Equipment	O
Identity	O
(	O
IMEI	O
)	O
Integrated	O
Circuit	O
Card	O
ID	O
(	O
ICCID	O
)	O
Mobile	O
equipment	O
identifier	O
(	O
MEID	O
)	O
Version	O
number	O
Product	O
number	O
The	O
profile	O
installations	O
differ	O
depending	O
on	O
the	O
iOS	B-System
.	O

For	O
versions	O
11.0	O
and	O
11.4	O
,	O
the	O
installation	O
is	O
straightforward	O
.	O

If	O
a	O
user	O
visits	O
the	O
profile	O
host	O
website	O
and	O
allows	O
the	O
installer	O
to	O
download	O
,	O
the	O
iOS	B-System
system	O
will	O
go	O
directly	O
to	O
the	O
“	O
Install	O
Profile	O
”	O
page	O
(	O
which	O
shows	O
a	O
verified	O
safety	O
certificate	O
)	O
,	O
and	O
then	O
request	O
the	O
users	O
’	O
passcode	O
for	O
the	O
last	O
step	O
of	O
installation	O
.	O

On	O
later	O
versions	O
,	O
specifically	O
iOS	B-System
12.1.1	I-System
and	O
iOS	B-System
12.2	I-System
,	O
the	O
process	O
is	O
different	O
.	O

After	O
the	O
profile	O
is	O
downloaded	O
,	O
the	O
iOS	B-System
system	O
will	O
first	O
ask	O
users	O
to	O
review	O
the	O
profile	O
in	O
their	O
settings	O
if	O
they	O
want	O
to	O
install	O
it	O
.	O

Users	O
can	O
see	O
a	O
“	O
Profile	O
Downloaded	O
”	O
added	O
in	O
their	O
settings	O
(	O
this	O
feature	O
is	O
in	O
iOS	B-System
12.2	I-System
,	O
but	O
not	O
on	O
iOS	B-System
12.1.1	I-System
)	O
.	O

This	O
gives	O
users	O
a	O
chance	O
to	O
see	O
details	O
and	O
better	O
understand	O
any	O
changes	O
made	O
.	O

After	O
the	O
review	O
,	O
the	O
process	O
is	O
the	O
same	O
as	O
above	O
.	O

After	O
the	O
profile	O
is	O
installed	O
,	O
the	O
user	O
will	O
then	O
be	O
redirected	O
to	O
another	O
Apple	B-Organization
phishing	O
site	O
.	O

The	O
phishing	O
site	O
uses	O
the	O
gathered	O
information	O
as	O
its	O
GET	O
parameter	O
,	O
allowing	O
the	O
attacker	O
to	O
access	O
the	O
stolen	O
information	O
.	O

Ongoing	O
activity	O
While	O
monitoring	O
this	O
particular	O
threat	O
,	O
we	O
found	O
another	O
XLoader	B-Malware
variant	O
posing	O
as	O
a	O
pornography	O
app	O
aimed	O
at	O
South	O
Korean	O
users	O
.	O

The	O
"	O
porn	O
kr	O
sex	O
''	O
APK	O
connects	O
to	O
a	O
malicious	O
website	O
that	O
runs	O
XLoader	B-Malware
in	O
the	O
background	O
.	O

The	O
website	O
uses	O
a	O
different	O
fixed	O
twitter	B-Organization
account	O
(	O
https	B-Indicator
:	I-Indicator
//twitter.com/fdgoer343	I-Indicator
)	O
.	O

This	O
attack	O
,	O
however	O
,	O
seems	O
exclusive	O
to	O
Android	B-System
users	O
,	O
as	O
it	O
does	O
not	O
have	O
the	O
code	O
to	O
attack	O
iOS	B-System
devices	O
.	O

Succeeding	O
monitoring	O
efforts	O
revealed	O
a	O
newer	O
variant	O
that	O
exploits	O
the	O
social	O
media	O
platforms	O
Instagram	B-Organization
and	O
Tumblr	B-Organization
instead	O
of	O
Twitter	B-Organization
to	O
hide	O
its	O
C	O
&	O
C	O
address	O
.	O

We	O
labeled	O
this	O
new	O
variant	O
XLoader	B-Malware
version	O
7.0	O
,	O
because	O
of	O
the	O
different	O
deployment	O
method	O
and	O
its	O
use	O
of	O
the	O
native	O
code	O
to	O
load	O
the	O
payload	O
and	O
hide	O
in	O
Instagram	B-Organization
and	O
Tumblr	B-Organization
profiles	O
.	O

These	O
more	O
recent	O
developments	O
indicate	O
that	O
XLoader	B-Malware
is	O
still	O
evolving	O
.	O

Adding	O
connections	O
to	O
FakeSpy	B-Malware
We	O
have	O
been	O
seeing	O
activity	O
from	O
XLoader	B-Malware
since	O
2018	O
,	O
and	O
have	O
since	O
followed	O
up	O
our	O
initial	O
findings	O
with	O
a	O
detailed	O
research	O
revealing	O
a	O
wealth	O
of	O
activity	O
dating	O
back	O
to	O
as	O
early	O
as	O
January	O
2015	O
,	O
which	O
outlined	O
a	O
major	O
discovery—its	O
connection	O
to	O
FakeSpy	B-Malware
.	O

The	O
emergence	O
of	O
XLoader	B-Malware
6.0	I-Malware
does	O
not	O
only	O
indicate	O
that	O
the	O
threat	O
actors	O
behind	O
it	O
remain	O
active	O
;	O
it	O
also	O
holds	O
fresh	O
evidence	O
of	O
its	O
connection	O
to	O
FakeSpy	B-Malware
.	O

One	O
such	O
immediately	O
apparent	O
connection	O
was	O
the	O
similar	O
deployment	O
technique	O
used	O
by	O
both	O
XLoader	B-Malware
6.0	I-Malware
and	O
FakeSpy	B-Malware
.	O

It	O
had	O
again	O
cloned	O
a	O
different	O
legitimate	O
Japanese	O
website	O
to	O
host	O
its	O
malicious	O
app	O
,	O
similar	O
to	O
what	O
FakeSpy	B-Malware
had	O
also	O
done	O
before	O
.	O

Their	O
similarity	O
is	O
made	O
more	O
apparent	O
by	O
looking	O
at	O
their	O
naming	O
method	O
for	O
downloadable	O
files	O
,	O
domain	O
structure	O
of	O
fake	O
websites	O
and	O
other	O
details	O
of	O
their	O
deployment	O
techniques	O
,	O
exemplified	O
in	O
figure	O
10	O
.	O

XLoader	B-Malware
6.0	I-Malware
also	O
mirrors	O
the	O
way	O
FakeSpy	B-Malware
hides	O
its	O
real	O
C	O
&	O
C	O
server	O
.	O

When	O
before	O
it	O
had	O
used	O
several	O
different	O
social	O
media	O
platforms	O
,	O
it	O
now	O
uses	O
the	O
Twitter	B-Organization
platform	O
,	O
something	O
FakeSpy	B-Malware
has	O
done	O
in	O
its	O
past	O
attacks	O
.	O

Analysis	O
of	O
the	O
malicious	O
iOS	B-System
profile	O
also	O
revealed	O
further	O
connections	O
,	O
as	O
the	O
profile	O
can	O
also	O
be	O
downloaded	O
from	O
a	O
website	O
that	O
FakeSpy	B-Malware
deployed	O
early	O
this	O
year	O
.	O

Conclusion	O
and	O
security	O
recommendations	O
The	O
continued	O
monitoring	O
of	O
XLoader	B-Malware
showed	O
how	O
its	O
operators	O
continuously	O
changed	O
its	O
features	O
,	O
such	O
as	O
its	O
attack	O
vector	O
deployment	O
infrastructure	O
and	O
deployment	O
techniques	O
.	O

This	O
newest	O
entry	O
seems	O
to	O
indicate	O
that	O
these	O
changes	O
won	O
’	O
t	O
be	O
stopping	O
soon	O
.	O

Being	O
aware	O
of	O
this	O
fact	O
can	O
help	O
create	O
defensive	O
strategies	O
,	O
as	O
well	O
as	O
prepare	O
for	O
upcoming	O
attacks	O
.	O

In	O
addition	O
,	O
just	O
as	O
uncovering	O
new	O
characteristics	O
is	O
important	O
,	O
finding	O
ones	O
we	O
’	O
ve	O
also	O
seen	O
in	O
a	O
different	O
malware	O
family	O
like	O
FakeSpy	B-Malware
also	O
provides	O
valuable	O
insight	O
.	O

Links	O
between	O
XLoader	B-Malware
and	O
FakeSpy	B-Malware
can	O
give	O
clues	O
to	O
the	O
much	O
broader	O
inner	O
workings	O
of	O
the	O
threat	O
actors	O
behind	O
them	O
.	O

Perhaps	O
more	O
information	O
on	O
XLoader	B-Malware
will	O
be	O
known	O
in	O
the	O
future	O
.	O

For	O
now	O
,	O
users	O
can	O
make	O
the	O
best	O
of	O
the	O
knowledge	O
they	O
have	O
now	O
to	O
significantly	O
reduce	O
the	O
effectivity	O
of	O
such	O
malware	O
.	O

Users	O
of	O
iOS	B-System
can	O
remove	O
the	O
malicious	O
profile	O
using	O
the	O
Apple	B-Organization
Configurator	O
2	O
,	O
Apple	B-Organization
’	O
s	O
official	O
iOS	B-System
helper	O
app	O
for	O
managing	O
Apple	B-Organization
devices	O
.	O

Following	O
simple	O
best	O
practices	O
,	O
like	O
strictly	O
downloading	O
applications	O
or	O
any	O
files	O
from	O
trusted	O
sources	O
and	O
being	O
wary	O
of	O
unsolicited	O
messages	O
,	O
can	O
also	O
prevent	O
similar	O
attacks	O
from	O
compromising	O
devices	O
.	O

Indicators	O
of	O
Compromise	O
SHA256	O
Package	O
App	O
label	O
332e68d865009d627343b89a5744843e3fde4ae870193f36b82980363439a425	B-Indicator
ufD.wykyx.vlhvh	B-Indicator
SEX	O
kr	O
porn	O
403401aa71df1830d294b78de0e5e867ee3738568369c48ffafe1b15f3145588	B-Indicator
ufD.wyjyx.vahvh	B-Indicator
佐川急便	O
466dafa82a4460dcad722d2ad9b8ca332e9a896fc59f06e16ebe981ad3838a6b	B-Indicator

com.dhp.ozqh	B-Indicator
Facebook	B-Organization
5022495104c280286e65184e3164f3f248356d065ad76acef48ee2ce244ffdc8	B-Indicator
ufD.wyjyx.vahvh	B-Indicator
Anshin	O
Scan	O
a0f3df39d20c4eaa410a61a527507dbc6b17c7f974f76e13181e98225bda0511	B-Indicator
com.aqyh.xolo	B-Indicator
佐川急便	O
cb412b9a26c1e51ece7a0e6f98f085e1c27aa0251172bf0a361eb5d1165307f7	B-Indicator

jp.co.sagawa.SagawaOfficialApp	B-Indicator
佐川急便	O
Malicious	O
URLs	O
:	O
hxxp	B-Indicator
:	I-Indicator
//38	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
27	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
99	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
11/xvideo/	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//apple-icloud	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
qwe-japan	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//apple-icloud	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
qwq-japan	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com/	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//apple-icloud	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
zqo-japan	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com/	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//files.spamo	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
jp/佐川急便.apk	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-qae	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-qaf	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-qau	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-qaw	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-wqe	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-wqo	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-wqp	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-wqq	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-wqu	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//mailsa-wqw	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//nttdocomo-qae	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//nttdocomo-qaq	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//nttdocomo-qaq	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com/aa	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//nttdocomo-qar	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//nttdocomo-qat	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//nttdocomo-qaw	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//sagawa-reg	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com/	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//www	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
711231	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//www	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
759383	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//www	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
923525	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//www	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
923915	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//www	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
975685	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
Malicious	O
Twitter	B-Organization
accounts	O
:	O
https	B-Indicator
:	I-Indicator
//twitter.com/lucky88755	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/lucky98745	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/lucky876543	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/luckyone1232	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/sadwqewqeqw	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/gyugyu87418490	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/fdgoer343	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/sdfghuio342	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/asdqweqweqeqw	I-Indicator
https	B-Indicator
:	I-Indicator
//twitter.com/ukenivor3	I-Indicator

Malicious	O
Instagram	B-Organization
account	O
:	O
https	B-Indicator
:	I-Indicator
//www.instagram.com/freedomguidepeople1830/	I-Indicator
Malicious	O
Tumblr	B-Organization
accounts	O
:	O
https	B-Indicator
:	I-Indicator
//mainsheetgyam.tumblr.com/	I-Indicator
https	B-Indicator
:	I-Indicator
//hormonaljgrj.tumblr.com/	I-Indicator
https	B-Indicator
:	I-Indicator
//globalanab.tumblr.com/	I-Indicator
C	O
&	O
C	O
addresses	O
:	O
104	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
160	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
191	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
190:8822	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
204	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
87:28833	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
204	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
87:28844	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
204	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
87:28855	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
205	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
122:28833	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
205	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
122:28844	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
205	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
122:28855	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
205	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
132:28833	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
205	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
132:28844	I-Indicator
61	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
230	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
205	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
132:28855	I-Indicator
GoldenCup	B-Malware
:	O
New	O
Cyber	O
Threat	O
Targeting	O
World	O
Cup	O
Fans	O
As	O
the	O
World	O
Cup	O
launches	O
,	O
so	O
does	O
a	O
new	O
threat	O
Officials	O
from	O
the	O
Israeli	B-Organization
Defense	I-Organization
Force	I-Organization
recently	O
uncovered	O
an	O
Android	B-System
Spyware	O
campaign	O
targeting	O
Israeli	O
soldiers	O
and	O
orchestrated	O
by	O
"	O
Hamas	B-Organization
.	O

''	O
The	O
latest	O
samples	O
attributed	O
to	O
this	O
campaign	O
were	O
discovered	O
by	O
security	O
researchers	O
from	O
ClearSky	B-Organization
.	O

In	O
our	O
research	O
,	O
we	O
focus	O
on	O
the	O
most	O
recent	O
sample	O
,	O
an	O
application	O
dubbed	O
as	O
"	O
Golden	B-Malware
Cup	I-Malware
''	O
,	O
launched	O
just	O
before	O
the	O
start	O
of	O
World	O
Cup	O
2018	O
.	O

Distribution	O
/	O
Infection	O
When	O
this	O
campaign	O
started	O
at	O
the	O
start	O
of	O
2018	O
,	O
the	O
malware	O
(	O
"	O
GlanceLove	B-Malware
''	O
,	O
"	O
WinkChat	B-Malware
''	O
)	O
was	O
distributed	O
by	O
the	O
perpetrators	O
mainly	O
via	O
fake	O
Facebook	B-System
profiles	O
,	O
attempting	O
to	O
seduce	O
IDF	O
soldiers	O
to	O
socialize	O
on	O
a	O
different	O
platform	O
(	O
their	O
malware	O
)	O
.	O

As	O
this	O
approach	O
was	O
not	O
a	O
great	O
success	O
,	O
their	O
last	O
attempt	O
was	O
to	O
quickly	O
create	O
a	O
World	O
Cup	O
app	O
and	O
this	O
time	O
distribute	O
it	O
to	O
Israeli	O
citizens	O
,	O
not	O
just	O
soldiers	O
.	O

The	O
official	O
“	O
Golden	B-Malware
Cup	I-Malware
”	O
Facebook	B-System
page	O
.	O

The	O
short	O
URL	O
redirects	O
to	O
the	O
application	O
page	O
at	O
Google	B-System
Play	I-System
.	O

The	O
official	O
“	O
Golden	B-Malware
Cup	I-Malware
”	O
Facebook	B-System
page	O
.	O

The	O
short	O
URL	O
redirects	O
to	O
the	O
application	O
page	O
at	O
Google	B-System
Play	I-System
.	O

We	O
assume	O
it	O
was	O
rushed	O
because	O
,	O
unlike	O
GlanceLove	B-Malware
,	O
it	O
lacked	O
any	O
real	O
obfuscation	O
.	O

Even	O
the	O
C	O
&	O
C	O
server	O
side	O
was	O
mostly	O
exposed	O
with	O
the	O
file	O
listing	O
available	O
for	O
everyone	O
to	O
traverse	O
through	O
it	O
.	O

It	O
contained	O
approximately	O
8GB	O
of	O
stolen	O
data	O
.	O

A	O
recent	O
whois	O
of	O
“	O
goldncup.com	B-Indicator
”	O
.	O

Creation	O
date	O
is	O
a	O
week	O
before	O
the	O
start	O
of	O
the	O
tournament	O
.	O

A	O
recent	O
whois	O
of	O
“	O
goldncup.com	B-Indicator
”	O
.	O

Creation	O
date	O
is	O
a	O
week	O
before	O
the	O
start	O
of	O
the	O
tournament	O
.	O

How	O
it	O
Works	O
In	O
order	O
to	O
get	O
into	O
the	O
Google	B-System
Play	I-System
Store	O
,	O
the	O
malware	O
uses	O
a	O
phased	O
approach	O
which	O
is	O
quite	O
a	O
common	O
practice	O
for	O
malware	O
authors	O
these	O
days	O
.	O

The	O
original	O
app	O
looks	O
innocent	O
,	O
with	O
most	O
of	O
its	O
code	O
aimed	O
at	O
implementing	O
the	O
real	O
features	O
that	O
the	O
app	O
claims	O
to	O
provide	O
.	O

In	O
addition	O
,	O
it	O
collects	O
identifiers	O
and	O
some	O
data	O
from	O
the	O
device	O
.	O

After	O
getting	O
a	O
command	O
from	O
the	O
C	O
&	O
C	O
,	O
the	O
app	O
is	O
able	O
to	O
download	O
a	O
malicious	O
payload	O
in	O
the	O
form	O
of	O
a	O
.dex	O
file	O
that	O
is	O
being	O
dynamically	O
loaded	O
adding	O
the	O
additional	O
malicious	O
capabilities	O
.	O

In	O
this	O
way	O
,	O
the	O
malware	O
authors	O
can	O
submit	O
their	O
app	O
and	O
add	O
the	O
malicious	O
capabilities	O
only	O
after	O
their	O
app	O
is	O
live	O
on	O
the	O
Play	B-System
Store	I-System
.	O

Communication	O
with	O
the	O
C	O
&	O
C	O
In	O
order	O
to	O
communicate	O
with	O
its	O
C	O
&	O
C	O
,	O
the	O
app	O
uses	O
the	O
MQTT	O
(	O
Message	O
Queuing	O
Telemetry	O
Transport	O
)	O
protocol	O
,	O
which	O
is	O
transported	O
over	O
TCP	B-Indicator
port	I-Indicator
1883	I-Indicator
.	O

Initiating	O
the	O
MQTT	O
client	O
.	O

Initiating	O
the	O
MQTT	O
client	O
.	O

Initiating	O
the	O
MQTT	O
client	O
.	O

The	O
app	O
connects	O
to	O
the	O
MQTT	O
broker	O
with	O
hardcoded	O
username	O
and	O
password	O
and	O
a	O
unique	O
device	O
identifier	O
generated	O
for	O
each	O
device	O
.	O

The	O
MQTT	O
connection	O
to	O
broker	O
The	O
MQTT	O
connection	O
to	O
broker	O
The	O
MQTT	O
communication	O
is	O
used	O
primarily	O
to	O
update	O
the	O
device	O
state	O
and	O
get	O
commands	O
from	O
the	O
C	O
&	O
C	O
.	O

It	O
uses	O
different	O
topics	O
that	O
include	O
the	O
unique	O
device	O
identifier	O
,	O
which	O
side	O
is	O
sending	O
the	O
message	O
,	O
and	O
whether	O
it	O
is	O
information	O
message	O
or	O
command	O
.	O

HTTP	O
Communication	O
In	O
addition	O
to	O
the	O
MQTT	O
communication	O
,	O
the	O
app	O
also	O
uses	O
plain	O
text	O
HTTP	O
communication	O
in	O
order	O
to	O
download	O
the	O
.dex	O
file	O
and	O
upload	O
collected	O
data	O
.	O

All	O
of	O
the	O
files	O
that	O
are	O
being	O
uploaded	O
or	O
downloaded	O
are	O
zip	O
files	O
encrypted	O
by	O
AES	O
with	O
ECB	O
mode	O
.	O

The	O
key	O
for	O
each	O
file	O
is	O
generated	O
randomly	O
and	O
stored	O
in	O
the	O
encrypted	O
file	O
with	O
a	O
fixed	O
offset	O
.	O

In	O
order	O
to	O
upload	O
the	O
file	O
,	O
the	O
app	O
uses	O
a	O
basic	O
REST	O
communication	O
with	O
the	O
server	O
,	O
checking	O
if	O
the	O
file	O
exists	O
and	O
uploading	O
it	O
if	O
it	O
isn	O
’	O
t	O
.	O

The	O
path	O
that	O
is	O
used	O
for	O
the	O
uploads	O
is	O
:	O
http	B-Indicator
:	I-Indicator
//	I-Indicator
/apps/d/p/op.php	I-Indicator
The	O
communication	O
looks	O
like	O
this	O
:	O
First	O
Phase	O
The	O
first	O
phase	O
of	O
the	O
app	O
’	O
s	O
attack	O
flow	O
collects	O
device	O
information	O
and	O
a	O
list	O
of	O
apps	O
installed	O
on	O
the	O
device	O
.	O

These	O
are	O
then	O
uploaded	O
to	O
the	O
C	O
&	O
C	O
HTTP	O
server	O
.	O

The	O
collection	O
of	O
basic	O
device	O
information	O
.	O

The	O
collection	O
of	O
basic	O
device	O
information	O
.	O

In	O
addition	O
,	O
at	O
this	O
stage	O
the	O
app	O
can	O
process	O
one	O
of	O
these	O
commands	O
:	O
•	O
Collect	O
device	O
info	O
•	O
Install	O
app	O
•	O
Is	O
online	O
?	O

•	O
Change	O
server	O
domain	O
Out	O
of	O
these	O
,	O
the	O
most	O
interesting	O
command	O
is	O
the	O
“	O
install	O
app	O
”	O
command	O
that	O
downloads	O
an	O
encrypted	O
zip	O
file	O
containing	O
the	O
second	O
phase	O
dex	O
file	O
,	O
unpacks	O
and	O
loads	O
it	O
.	O

Second	O
Phase	O
The	O
second	O
phase	O
dex	O
file	O
contains	O
3	O
main	O
services	O
that	O
are	O
being	O
used	O
:	O
•	O
ConnManager	O
-	O
handles	O
connections	O
to	O
the	O
C	O
&	O
C	O
•	O
ReceiverManager	O
-	O
waits	O
for	O
incoming	O
calls	O
/	O
app	O
installations	O
•	O
TaskManager	O
-	O
manages	O
the	O
data	O
collection	O
tasks	O
The	O
C	O
&	O
C	O
server	O
address	O
is	O
different	O
than	O
the	O
one	O
that	O
is	O
used	O
by	O
the	O
first	O
phase	O
,	O
so	O
the	O
app	O
reconnects	O
to	O
the	O
new	O
server	O
as	O
well	O
as	O
starts	O
the	O
periodic	O
data	O
collector	O
tasks	O
.	O

By	O
analyzing	O
the	O
TaskManager	O
class	O
we	O
can	O
see	O
the	O
new	O
commands	O
that	O
are	O
supported	O
at	O
this	O
stage	O
:	O
As	O
can	O
be	O
seen	O
in	O
the	O
code	O
snippet	O
above	O
,	O
there	O
are	O
quite	O
a	O
lot	O
of	O
data	O
collection	O
tasks	O
that	O
are	O
now	O
available	O
:	O
Collect	O
device	O
info	O
Track	O
location	O
Upload	O
contacts	O
information	O
Upload	O
sent	O
and	O
received	O
SMS	O
messages	O
Upload	O
images	O
Upload	O
video	O
files	O
Send	O
recursive	O
dirlist	O
of	O
the	O
external	O
storage	O
Upload	O
specific	O
files	O
Record	O
audio	O
using	O
the	O
microphone	O
Record	O
calls	O
Use	O
the	O
camera	O
to	O
capture	O
bursts	O
of	O
snapshots	O
Those	O
tasks	O
can	O
either	O
run	O
periodically	O
,	O
on	O
event	O
(	O
such	O
as	O
incoming	O
call	O
)	O
or	O
when	O
getting	O

a	O
command	O
from	O
the	O
C	O
&	O
C	O
server	O
.	O

Mitigations	O
Stay	O
protected	O
from	O
mobile	O
malware	O
by	O
taking	O
these	O
precautions	O
:	O
Do	O
not	O
download	O
apps	O
from	O
unfamiliar	O
sites	O
Only	O
install	O
apps	O
from	O
trusted	O
sources	O
Pay	O
close	O
attention	O
to	O
the	O
permissions	O
requested	O
by	O
apps	O
Install	O
a	O
suitable	O
mobile	O
security	O
app	O
,	O
such	O
as	O
SEP	O
Mobile	O
or	O
Norton	O
,	O
to	O
protect	O
your	O
device	O
and	O
data	O
Keep	O
your	O
operating	O
system	O
up	O
to	O
date	O
Make	O
frequent	O
backups	O
of	O
important	O
data	O
Indicators	O
of	O
Compromise	O
(	O
IoCs	O
)	O
Package	O
names	O
:	O
anew.football.cup.world.com.worldcup	B-Indicator
com.coder.glancelove	B-Indicator
com.winkchat	I-Indicator
APK	O
SHA2	O
:	O
166f3a863bb2b66bda9c76dccf9529d5237f6394721f46635b053870eb2fcc5a	B-Indicator

b45defca452a640b303288131eb64c485f442aae0682a3c56489d24d59439b47	B-Indicator
d9601735d674a9e55546fde0bffde235bc5f2546504b31799d874e8c31d5b6e9	B-Indicator
2ce54d93510126fca83031f9521e40cd8460ae564d3d927e17bd63fb4cb20edc	B-Indicator
67b1a1e7b505ac510322b9d4f4fc1e8a569d6d644582b588faccfeeaa4922cb7	B-Indicator

1664cb343ee830fa94725fed143b119f7e2351307ed0ce04724b23469b9002f2	B-Indicator
Loaded	O
DEX	O
SHA2	O
:	O
afaf446a337bf93301b1d72855ccdd76112595f6e4369d977bea6f9721edf37e	B-Indicator
Domain/IP	O
:	O
goldncup	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
glancelove	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
autoandroidup	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
website	I-Indicator
mobilestoreupdate	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
website	I-Indicator
updatemobapp	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
website	I-Indicator
107	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
175	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
144	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
26	I-Indicator
192	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
64	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
114	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
147	I-Indicator
Red	B-Malware
Alert	I-Malware
2.0	I-Malware
:	O
Android	B-System
Trojan	O
targets	O
security-seekers	O
A	O
malicious	O
,	O
counterfeit	O
version	O
of	O
a	O
VPN	B-System
client	O
for	O
mobile	O
devices	O
targets	O
security-minded	O
victims	O
with	O
a	O
RAT	O
.	O

Written	O
by	O
Jagadeesh	O
Chandraiah	O
JULY	O
23	O
,	O
2018	O
SophosLabs	B-Organization
has	O
uncovered	O
a	O
mobile	O
malware	O
distribution	O
campaign	O
that	O
uses	O
advertising	O
placement	O
to	O
distribute	O
the	O
Red	B-Malware
Alert	I-Malware
Trojan	I-Malware
,	O
linking	O
counterfeit	O
branding	O
of	O
well-known	O
apps	O
to	O
Web	O
pages	O
that	O
deliver	O
an	O
updated	O
,	O
2.0	O
version	O
of	O
this	O
bank	O
credential	O
thief	O
.	O

The	O
group	O
distributing	O
this	O
family	O
of	O
malware	O
decorates	O
it	O
in	O
the	O
branding	O
and	O
logos	O
of	O
well-known	O
social	O
media	O
or	O
media	O
player	O
apps	O
,	O
system	O
update	O
patches	O
,	O
or	O
(	O
in	O
its	O
most	O
recent	O
campaign	O
)	O
VPN	B-System
client	O
apps	O
in	O
an	O
attempt	O
to	O
lure	O
users	O
into	O
downloading	O
,	O
installing	O
,	O
and	O
elevating	O
the	O
privileges	O
of	O
a	O
Trojanized	O
app	O
hosted	O
on	O
a	O
site	O
not	O
affiliated	O
with	O
any	O
reputable	O
app	O
market	O
or	O
store	O
.	O

Aside	O
from	O
the	O
inescapable	O
irony	O
of	O
disguising	O
a	O
security-reducing	O
Trojan	O
as	O
an	O
ostensibly	O
security-enhancing	O
app	O
,	O
and	O
the	O
righteous	O
affront	O
to	O
the	O
whole	O
concept	O
of	O
a	O
VPN	O
’	O
s	O
purpose	O
a	O
Trojan	O
so	O
disguised	O
inspires	O
,	O
this	O
represents	O
an	O
escalation	O
in	O
the	O
variety	O
of	O
app	O
types	O
targeted	O
by	O
this	O
campaign	O
of	O
bankbots	O
in	O
disguise	O
.	O

Red	B-Malware
Alert	I-Malware
Plays	O
Dress-Up	O
In	O
the	O
wild	O
,	O
we	O
found	O
Web	O
pages	O
designed	O
to	O
(	O
vaguely	O
)	O
resemble	O
legitimate	O
app	O
market	O
pages	O
,	O
hosting	O
files	O
for	O
download	O
that	O
have	O
been	O
disguised	O
as	O
a	O
legitimate	O
mobile	O
application	O
of	O
moderately	O
broad	O
appeal	O
,	O
such	O
as	O
a	O
media	O
player	O
or	O
social	O
media	O
app	O
.	O

But	O
the	O
categories	O
targeted	O
by	O
this	O
group	O
seem	O
to	O
be	O
broadening	O
with	O
the	O
inclusion	O
of	O
VPN	B-System
software	O
.	O

The	O
Web	O
page	O
shown	O
here	O
on	O
the	O
left	O
is	O
hosted	O
on	O
a	O
domain	O
that	O
seems	O
apt	O
:	O
free-vpn	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
download	I-Indicator
.	O

Investigation	O
of	O
this	O
domain	O
led	O
to	O
additional	O
domains	O
that	O
appear	O
to	O
have	O
been	O
registered	O
for	O
use	O
with	O
the	O
campaign	O
,	O
but	O
are	O
not	O
in	O
use	O
yet	O
.	O

(	O
You	O
can	O
find	O
additional	O
IoCs	O
at	O
the	O
end	O
of	O
this	O
article	O
)	O
As	O
you	O
can	O
see	O
,	O
the	O
Web	O
page	O
uses	O
a	O
similar	O
colour	O
scheme	O
as	O
,	O
and	O
the	O
icon	O
design	O
from	O
,	O
a	O
legitimate	O
VPN	O
application	O
(	O
VPN	O
Proxy	O
Master	O
)	O
found	O
on	O
the	O
Google	B-System
Play	I-System
store	I-System
.	O

The	O
fake	O
doesn	O
’	O
t	O
quite	O
nail	O
the	O
app	O
name	O
.	O

In	O
addition	O
to	O
“	O
Free	B-System
VPN	I-System
Master	I-System
Android	I-System
,	O
”	O
we	O
’	O
ve	O
observed	O
Red	B-Malware
Alert	I-Malware
2.0	I-Malware
Trojans	O
in	O
the	O
wild	O
disguising	O
themselves	O
using	O
names	O
like	O
:	O
Flash	B-System
Player	I-System
or	O
Update	B-System
Flash	I-System
Player	I-System
Android	B-System
Update	I-System
or	O
Android	B-System
Antivirus	I-System
Chrome	B-System
Update	I-System
or	O
Google	B-System
Update	I-System
Update	B-System
Google	I-System
Market	I-System
WhatsApp	B-System
Viber	B-System
OneCoin	B-System
Wallet	B-System
Pornhub	O
Tactic	O
FlashLight	O
or	O
PROFlashLight	O
Finanzonline	O
The	O
vast	O
majority	O
of	O
in-the-wild	O
Red	B-Malware
Alert	I-Malware
2.0	I-Malware
samples	I-Malware
falsely	O
present	O
themselves	O
as	O
Adobe	B-System
Flash	I-System
player	I-System
for	O
Android	B-System
,	O
a	O
utility	O
that	O
Adobe	B-Organization
stopped	O
supporting	O
years	O
ago	O
.	O

Our	O
logs	O
show	O
a	O
number	O
of	O
simultaneous	B-Malware
Red	I-Malware
Alert	I-Malware
2.0	I-Malware
campaigns	I-Malware
in	O
operation	O
,	O
many	O
(	O
but	O
not	O
all	O
)	O
hosted	O
on	O
dynamic	O
DNS	O
domains	O
.	O

The	O
Red	B-Malware
Alert	I-Malware
Payload	I-Malware
Once	O
installed	O
,	O
the	O
malware	O
requests	O
Device	O
Administrator	O
privileges	O
.	O

If	O
the	O
malware	O
obtains	O
device	O
administrator	O
rights	O
,	O
it	O
will	O
be	O
able	O
to	O
lock	O
the	O
screen	O
by	O
itself	O
,	O
expire	O
the	O
password	O
,	O
and	O
resist	O
being	O
uninstalled	O
through	O
normal	O
methods	O
.	O

Device	O
admin	O
request	O
from	O
app	O
that	O
says	O
it	O
is	O
WhatsApp	O
The	O
app	O
then	O
stays	O
in	O
the	O
background	O
listening	O
to	O
commands	O
from	O
the	O
cybercrooks	O
.	O

Within	O
some	O
of	O
the	O
first	O
of	O
those	O
commands	O
,	O
the	O
bot	O
typically	O
receives	O
a	O
list	O
of	O
banks	O
it	O
will	O
target	O
.	O

The	O
Trojan	O
works	O
by	O
creating	O
an	O
overlay	O
whenever	O
the	O
user	O
launches	O
the	O
banking	O
application	O
.	O

Currently	O
Running	O
Applications	O
Banking	O
Trojans	O
that	O
rely	O
on	O
the	O
overlay	O
mechanism	O
to	O
steal	O
information	O
need	O
to	O
know	O
what	O
application	O
is	O
in	O
the	O
foreground	O
.	O

They	O
do	O
this	O
not	O
only	O
to	O
identify	O
whether	O
the	O
use	O
of	O
a	O
particular	O
app	O
may	O
permit	O
them	O
to	O
harvest	O
another	O
credential	O
,	O
but	O
also	O
because	O
each	O
targeted	O
app	O
needs	O
to	O
have	O
an	O
overlay	O
mapped	O
to	O
its	O
design	O
,	O
so	O
the	O
Trojan	O
can	O
intercept	O
and	O
steal	O
user	O
data	O
.	O

This	O
quest	O
to	O
determine	O
the	O
currently	O
running	O
application	O
is	O
a	O
hallmark	O
of	O
overlay	O
malware	O
,	O
so	O
we	O
thought	O
we	O
’	O
d	O
take	O
a	O
closer	O
look	O
at	O
how	O
it	O
’	O
s	O
done	O
.	O

To	O
prevent	O
this	O
,	O
Android	B-System
’	O
s	O
engineers	O
regularly	O
release	O
updates	O
that	O
contain	O
bug	O
fixes	O
designed	O
to	O
prevent	O
apps	O
from	O
getting	O
the	O
list	O
of	O
currently	O
running	O
apps	O
without	O
explicit	O
permission	O
.	O

With	O
every	O
Android	B-System
update	O
,	O
the	O
malware	O
authors	O
are	O
forced	O
to	O
come	O
up	O
with	O
new	O
tricks	O
.	O

This	O
particular	O
case	O
is	O
not	O
an	O
exception	O
.	O

The	O
author	O
(	O
s	O
)	O
of	O
this	O
malware	O
wrote	O
separate	O
subroutines	O
that	O
identify	O
the	O
operating	O
system	O
version	O
and	O
fire	O
off	O
methods	O
to	O
obtain	O
a	O
list	O
of	O
currently	O
running	O
applications	O
known	O
to	O
work	O
on	O
that	O
particular	O
version	O
of	O
Android	B-System
.	O

First	O
,	O
they	O
use	O
the	O
built-in	O
toolbox	O
commands	O
to	O
determine	O
what	O
apps	O
are	O
running	O
.	O

If	O
that	O
doesn	O
’	O
t	O
work	O
,	O
they	O
try	O
to	O
use	O
queryUsageStats	O
:	O
When	O
the	O
malware	O
invokes	O
queryUsageStats	O
,	O
it	O
asks	O
for	O
the	O
list	O
of	O
applications	O
that	O
ran	O
in	O
the	O
last	O
1	O
million	O
milliseconds	O
(	O
16	O
minutes	O
and	O
40	O
seconds	O
)	O
.	O

String	O
Resources	O
Used	O
to	O
Store	O
App	O
Data	O
Red	B-Malware
Alert	I-Malware
2.0	I-Malware
stores	O
its	O
data	O
in	O
an	O
atypical	O
location	O
(	O
inside	O
the	O
Strings.xml	B-Indicator
file	I-Indicator
embedded	O
in	O
the	O
app	O
)	O
to	O
fetch	O
its	O
critical	O
data	O
,	O
such	O
as	O
the	O
C2	O
address	O
.	O

The	O
com.dsufabunfzs.dowiflubs	O
strings	O
in	O
the	O
screenshot	O
above	O
refer	O
to	O
the	O
internal	O
name	O
this	O
particular	O
malware	O
was	O
given	O
,	O
which	O
in	O
this	O
case	O
was	O
randomized	O
into	O
alphabet	O
salad	O
.	O

It	O
’	O
s	O
been	O
SophosLabs	O
’	O
observation	O
that	O
Red	B-Malware
Alert	I-Malware
Trojans	I-Malware
usually	O
have	O
a	O
randomized	O
internal	O
name	O
like	O
this	O
.	O

The	O
strings	O
section	O
of	O
the	O
app	O
contains	O
embedded	O
command-and-control	O
IP	O
addresses	O
,	O
ports	O
,	O
and	O
domain	O
names	O
in	O
plaintext	O
.	O

It	O
is	O
an	O
invaluable	O
source	O
of	O
intelligence	O
about	O
a	O
given	O
campaign	O
..	O
The	O
following	O
snippet	O
shows	O
the	O
location	O
within	O
the	O
Trojan	O
where	O
it	O
uses	O
SQLite	O
database	O
commands	O
to	O
store	O
and	O
recall	O
command-and-control	O
addresses	O
:	O
Backdoor	O
Commands	O
The	O
Red	B-Malware
Alert	I-Malware
code	I-Malware
also	O
contains	O
an	O
embedded	O
list	O
of	O
commands	O
the	O
botmaster	O
can	O
send	O
to	O
the	O
bot	O
.	O

The	O
malware	O
can	O
execute	O
a	O
variety	O
of	O
arbitrary	O
commands	O
,	O
including	O
(	O
for	O
example	O
)	O
intercepting	O
or	O
sending	O
text	O
messages	O
without	O
the	O
user	O
’	O
s	O
knowledge	O
,	O
obtaining	O
a	O
copy	O
of	O
the	O
victim	O
’	O
s	O
Address	B-System
Book	I-System
,	O
or	O
call	O
or	O
text	O
message	O
logs	O
,	O
or	O
sending	O
phone	O
network	O
feature	O
codes	O
(	O
also	O
known	O
as	O
USSD	O
codes	O
)	O
.	O

C2	O
and	O
Targeted	O
Banks	O
As	O
described	O
earlier	O
,	O
the	O
C2	O
domain	O
is	O
kept	O
in	O
the	O
app	O
’	O
s	O
resources	O
.	O

During	O
the	O
app	O
execution	O
,	O
the	O
malware	O
contacts	O
C2	O
domain	O
for	O
further	O
instructions	O
.	O

Most	O
of	O
the	O
network	O
traffic	O
we	O
’	O
ve	O
observed	O
is	O
HTTP	B-Indicator
.	O

The	O
C2	O
address	O
,	O
as	O
stored	O
in	O
samples	O
we	O
’	O
ve	O
seen	O
,	O
comprise	O
both	O
an	O
IP	O
address	O
and	O
port	O
number	O
;	O
So	O
far	O
,	O
all	O
the	O
samples	O
we	O
’	O
ve	O
tested	O
attempted	O
to	O
contact	O
an	O
IP	O
address	O
on	O
port	B-Indicator
7878/tcp	I-Indicator
.	O

If	O
the	O
main	O
C2	O
domain	O
is	O
not	O
responsive	O
,	O
the	O
bot	O
fetches	O
a	O
backup	O
C2	O
domain	O
from	O
a	O
Twitter	B-Organization
account	O
.	O

Static	O
analysis	O
of	O
the	O
code	O
reveals	O
that	O
the	O
malware	O
downloads	O
the	O
overlay	O
template	O
to	O
use	O
against	O
any	O
of	O
the	O
bank	O
(	O
s	O
)	O
it	O
is	O
targeting	O
.	O

The	O
malware	O
also	O
sends	O
regular	O
telemetry	O
back	O
to	O
its	O
C2	O
server	O
about	O
the	O
infected	O
device	O
in	O
the	O
form	O
of	O
an	O
HTTP	B-Indicator
POST	O
to	O
its	O
C2	O
server	O
.	O

It	O
uses	O
the	O
base	O
Dalvik	O
User-Agent	O
string	O
for	O
the	O
device	O
it	O
’	O
s	O
running	O
on	O
.	O

The	O
content	O
of	O
the	O
HTTP	B-Indicator
POST	O
data	O
is	O
telemetry	O
data	O
in	O
a	O
json	O
format	O
about	O
the	O
device	O
the	O
malware	O
is	O
running	O
on	O
.	O

The	O
list	O
of	O
banks	O
targeted	O
by	O
Red	B-Malware
Alert	I-Malware
2.0	I-Malware
includes	O
NatWest	O
,	O
Barclays	B-Organization
,	O
Westpac	O
,	O
and	O
Citibank	O
.	O

Red	B-Malware
Alert	I-Malware
2.0	I-Malware
is	O
a	O
banking	O
bot	O
that	O
is	O
currently	O
very	O
active	O
online	O
,	O
and	O
presents	O
a	O
risk	O
to	O
Android	O
devices	O
.	O

We	O
expect	O
to	O
see	O
more	O
diversification	O
in	O
the	O
social	O
engineering	O
lures	O
this	O
threat	O
group	O
employs	O
as	O
time	O
goes	O
on	O
.	O

So	O
far	O
,	O
legitimate	O
app	O
stores	O
appear	O
to	O
be	O
this	O
malware	O
’	O
s	O
Achilles	O
heel	O
;	O
disabling	O
the	O
installation	O
of	O
third-party	O
apps	O
has	O
been	O
an	O
effective	O
prevention	O
measure	O
.	O

Stick	O
to	O
Google	B-System
Play	I-System
and	O
use	O
VPN	O
software	O
from	O
reputable	O
vendors	O
.	O

Sophos	B-Organization
detects	O
all	O
the	O
samples	O
of	O
this	O
Trojan	O
family	O
as	O
Andr/Banker-GWC	O
and	O
Andr/Spybot-A	O
.	O

In	O
the	O
wild	O
,	O
these	O
are	O
only	O
distributed	O
as	O
a	O
direct	O
download	O
from	O
unofficial	O
Web	O
pages	O
(	O
“	O
third-party	O
”	O
app	O
)	O
and	O
not	O
through	O
legitimate	O
app	O
stores	O
.	O

Red	B-Malware
Alert	I-Malware
2.0	I-Malware
IoCs	O
list	O
C2	O
addresses	O
103.239.30.126:7878	B-Indicator
146.185.241.29:7878	B-Indicator
146.185.241.42:7878	B-Indicator
185.126.200.3:7878	B-Indicator
185.126.200.12:7878	B-Indicator
185.126.200.15:7878	B-Indicator
185.126.200.18:7878	B-Indicator
185.165.28.15:7878	B-Indicator
185.243.243.241:7878	B-Indicator
185.243.243.244:7878	B-Indicator
185.243.243.245:7878	B-Indicator
Domains	O
Malware	O
source	O
Web	O
hosts	O

on	O
167.99.176.61	B-Indicator
:	O
free-androidvpn.date	B-Indicator
free-androidvpn.download	O
free-androidvpn.online	O
free-vpn.date	B-Indicator
free-vpn.download	O
free-vpn.online	O
Hashes	O
22fcfce096392f085218c3a78dd0fa4be9e67ed725bce42b965a27725f671cf	O
55292a4dde8727faad1c40c914cf1be9dfdcf4e67b515aa593bcd8d86e824372	B-Indicator

be92a751e5abbcd24151b509dbb4feb98ea46f367a99d6f86ed4a7c162461e31	B-Indicator
5c4d666cef84abc2a1ffd3b1060ef28fa3c6c3bb4fad1fa26db99350b41bea4c	B-Indicator
06081ab7faa729e33b9397a0e47548e75cbec3d43c50e6368e81d737552150a5	B-Indicator
753999cb19a4346042f973e30cf1158c44f2335ab65859d3bfa16bca4098e2ef	B-Indicator

As	O
a	O
result	O
of	O
a	O
lot	O
of	O
hard	O
work	O
done	O
by	O
our	O
security	O
research	O
teams	O
,	O
we	O
revealed	O
today	O
a	O
new	O
and	O
alarming	O
malware	O
campaign	O
.	O

The	O
attack	O
campaign	O
,	O
named	O
Gooligan	B-Malware
,	O
breached	O
the	O
security	O
of	O
over	O
one	O
million	O
Google	B-Organization
accounts	O
.	O

The	O
number	O
continues	O
to	O
rise	O
at	O
an	O
additional	O
13,000	O
breached	O
devices	O
each	O
day	O
.	O

Our	O
research	O
exposes	O
how	O
the	O
malware	O
roots	O
infected	O
devices	O
and	O
steals	O
authentication	O
tokens	O
that	O
can	O
be	O
used	O
to	O
access	O
data	O
from	O
Google	B-System
Play	I-System
,	O
Gmail	B-System
,	O
Google	B-System
Photos	I-System
,	O
Google	B-System
Docs	I-System
,	O
G	B-System
Suite	I-System
,	O
Google	B-System
Drive	I-System
,	O
and	O
more	O
.	O

Gooligan	B-Malware
is	O
a	O
new	O
variant	O
of	O
the	O
Android	O
malware	O
campaign	O
found	O
by	O
our	O
researchers	O
in	O
the	O
SnapPea	B-Malware
app	O
last	O
year	O
.	O

Check	B-Organization
Point	I-Organization
reached	O
out	O
to	O
the	O
Google	B-Organization
Security	I-Organization
team	O
immediately	O
with	O
information	O
on	O
this	O
campaign	O
.	O

Our	O
researchers	O
are	O
working	O
closely	O
with	O
Google	B-Organization
to	O
investigate	O
the	O
source	O
of	O
the	O
Gooligan	B-Malware
campaign	I-Malware
.	O

“	O
We	O
’	O
re	O
appreciative	O
of	O
both	O
Check	B-Organization
Point	I-Organization
’	O
s	O
research	O
and	O
their	O
partnership	O
as	O
we	O
’	O
ve	O
worked	O
together	O
to	O
understand	O
these	O
issues	O
,	O
”	O
said	O
Adrian	O
Ludwig	O
,	O
Google	B-Organization
’	O
s	O
director	O
of	O
Android	B-System
security	O
.	O

“	O
As	O
part	O
of	O
our	O
ongoing	O
efforts	O
to	O
protect	O
users	O
from	O
the	O
Ghost	B-Malware
Push	I-Malware
family	I-Malware
of	O
malware	O
,	O
we	O
’	O
ve	O
taken	O
numerous	O
steps	O
to	O
protect	O
our	O
users	O
and	O
improve	O
the	O
security	O
of	O
the	O
Android	B-System
ecosystem	O
overall.	O
”	O
We	O
are	O
very	O
encouraged	O
by	O
the	O
statement	O
Google	B-Organization
shared	O
with	O
us	O
addressing	O
the	O
issue	O
.	O

We	O
have	O
chosen	O
to	O
join	O
forces	O
to	O
continue	O
the	O
investigation	O
around	O
Gooligan	B-Malware
.	O

Google	B-Organization
also	O
stated	O
that	O
they	O
are	O
taking	O
numerous	O
steps	O
including	O
proactively	O
notifying	O
affected	O
accounts	O
,	O
revoking	O
affected	O
tokens	O
and	O
deploying	O
SafetyNet	O
improvements	O
to	O
protect	O
users	O
from	O
these	O
apps	O
in	O
the	O
future	O
.	O

Who	O
is	O
affected	O
?	O

Gooligan	B-Malware
potentially	O
affects	O
devices	O
on	O
Android	B-System
4	I-System
(	I-System
Jelly	I-System
Bean	I-System
,	I-System
KitKat	I-System
)	I-System
and	I-System
5	I-System
(	I-System
Lollipop	I-System
)	I-System
,	O
which	O
is	O
over	O
74	O
%	O
of	O
in-market	O
devices	O
today	O
.	O

About	O
57	O
%	O
of	O
these	O
devices	O
are	O
located	O
in	O
Asia	O
and	O
about	O
9	O
%	O
are	O
in	O
Europe	O
.	O

In	O
our	O
research	O
we	O
identified	O
tens	O
of	O
fake	O
applications	O
that	O
were	O
infected	O
with	O
this	O
malware	O
.	O

If	O
you	O
’	O
ve	O
downloaded	O
one	O
of	O
the	O
apps	O
listed	O
in	O
Appendix	O
A	O
,	O
below	O
,	O
you	O
might	O
be	O
infected	O
.	O

You	O
may	O
review	O
your	O
application	O
list	O
in	O
“	O
Settings	O
-	O
>	O
Apps	O
”	O
,	O
if	O
you	O
find	O
one	O
of	O
this	O
applications	O
,	O
please	O
consider	O
downloading	O
an	O
antivirus	O
product	O
such	O
as	O
Check	B-Organization
Point	I-Organization
ZoneAlarm	B-System
to	O
check	O
if	O
you	O
are	O
indeed	O
infected	O
.	O

We	O
have	O
noticed	O
that	O
hundreds	O
of	O
the	O
email	O
addresses	O
are	O
associated	O
with	O
enterprise	O
accounts	O
worldwide	O
.	O

How	O
do	O
you	O
know	O
if	O
your	O
Google	B-Organization
account	O
is	O
breached	O
?	O

You	O
can	O
check	O
if	O
your	O
account	O
is	O
compromised	O
by	O
accessing	O
the	O
following	O
web	O
site	O
that	O
we	O
created	O
:	O
https	B-Indicator
:	I-Indicator
//gooligan.checkpoint.com/	I-Indicator
.	O

If	O
your	O
account	O
has	O
been	O
breached	O
,	O
the	O
following	O
steps	O
are	O
required	O
:	O
A	O
clean	O
installation	O
of	O
an	O
operating	O
system	O
on	O
your	O
mobile	O
device	O
is	O
required	O
(	O
a	O
process	O
called	O
“	O
flashing	O
”	O
)	O
.	O

As	O
this	O
is	O
a	O
complex	O
process	O
,	O
we	O
recommend	O
powering	O
off	O
your	O
device	O
and	O
approaching	O
a	O
certified	O
technician	O
,	O
or	O
your	O
mobile	O
service	O
provider	O
,	O
to	O
request	O
that	O
your	O
device	O
be	O
“	O
re-flashed.	O
”	O
Change	O
your	O
Google	B-Organization
account	O
passwords	O
immediately	O
after	O
this	O
process	O
.	O

How	O
do	O
Android	O
devices	O
become	O
infected	O
?	O

We	O
found	O
traces	O
of	O
the	O
Gooligan	B-Malware
malware	O
code	O
in	O
dozens	O
of	O
legitimate-looking	O
apps	O
on	O
third-party	O
Android	B-System
app	O
stores	O
.	O

These	O
stores	O
are	O
an	O
attractive	O
alternative	O
to	O
Google	B-System
Play	I-System
because	O
many	O
of	O
their	O
apps	O
are	O
free	O
,	O
or	O
offer	O
free	O
versions	O
of	O
paid	O
apps	O
.	O

However	O
,	O
the	O
security	O
of	O
these	O
stores	O
and	O
the	O
apps	O
they	O
sell	O
aren	O
’	O
t	O
always	O
verified	O
.	O

Gooligan-infected	B-Malware
apps	O
can	O
also	O
be	O
installed	O
using	O
phishing	O
scams	O
where	O
attackers	O
broadcast	O
links	O
to	O
infected	O
apps	O
to	O
unsuspecting	O
users	O
via	O
SMS	O
or	O
other	O
messaging	O
services	O
.	O

How	O
did	O
Gooligan	B-Malware
emerge	O
?	O

Our	O
researchers	O
first	O
encountered	O
Gooligan	B-Malware
’	O
s	O
code	O
in	O
the	O
malicious	O
SnapPea	B-Malware
app	O
last	O
year	O
.	O

At	O
the	O
time	O
this	O
malware	O
was	O
reported	O
by	O
several	O
security	O
vendors	O
,	O
and	O
attributed	O
to	O
different	O
malware	O
families	O
like	O
Ghostpush	B-Malware
,	O
MonkeyTest	B-Malware
,	O
and	O
Xinyinhe	B-Malware
.	O

By	O
late	O
2015	O
,	O
the	O
malware	O
’	O
s	O
creators	O
had	O
gone	O
mostly	O
silent	O
until	O
the	O
summer	O
of	O
2016	O
when	O
the	O
malware	O
reappeared	O
with	O
a	O
more	O
complex	O
architecture	O
that	O
injects	O
malicious	O
code	O
into	O
Android	B-System
system	O
processes	O
.	O

The	O
change	O
in	O
the	O
way	O
the	O
malware	O
works	O
today	O
may	O
be	O
to	O
help	O
finance	O
the	O
campaign	O
through	O
fraudulent	O
ad	O
activity	O
.	O

The	O
malware	O
simulates	O
clicks	O
on	O
app	O
advertisements	O
provided	O
by	O
legitimate	O
ad	O
networks	O
and	O
forces	O
the	O
app	O
to	O
install	O
on	O
a	O
device	O
.	O

An	O
attacker	O
is	O
paid	O
by	O
the	O
network	O
when	O
one	O
of	O
these	O
apps	O
is	O
installed	O
successfully	O
.	O

Logs	O
collected	O
by	O
Check	B-Organization
Point	I-Organization
researchers	O
show	O
that	O
every	O
day	O
Gooligan	B-Malware
installs	O
at	O
least	O
30,000	O
apps	O
fraudulently	O
on	O
breached	O
devices	O
or	O
over	O
2	O
million	O
apps	O
since	O
the	O
campaign	O
began	O
.	O

How	O
does	O
Gooligan	B-Malware
work	O
?	O

The	O
infection	O
begins	O
when	O
a	O
user	O
downloads	O
and	O
installs	O
a	O
Gooligan-infected	B-Malware
app	O
on	O
a	O
vulnerable	O
Android	O
device	O
.	O

Our	O
research	O
team	O
has	O
found	O
infected	O
apps	O
on	O
third-party	O
app	O
stores	O
,	O
but	O
they	O
could	O
also	O
be	O
downloaded	O
by	O
Android	B-System
users	O
directly	O
by	O
tapping	O
malicious	O
links	O
in	O
phishing	O
attack	O
messages	O
.	O

After	O
an	O
infected	O
app	O
is	O
installed	O
,	O
it	O
sends	O
data	O
about	O
the	O
device	O
to	O
the	O
campaign	O
’	O
s	O
Command	O
and	O
Control	O
(	O
C	O
&	O
C	O
)	O
server	O
.	O

Gooligan	B-Malware
then	O
downloads	O
a	O
rootkit	O
from	O
the	O
C	O
&	O
C	O
server	O
that	O
takes	O
advantage	O
of	O
multiple	O
Android	B-System
4	I-System
and	I-System
5	I-System
exploits	O
including	O
the	O
well-known	O
VROOT	B-Vulnerability
(	O
CVE-2013-6282	B-Vulnerability
)	O
and	O
Towelroot	B-Vulnerability
(	O
CVE-2014-3153	B-Vulnerability
)	O
.	O

These	O
exploits	O
still	O
plague	O
many	O
devices	O
today	O
because	O
security	O
patches	O
that	O
fix	O
them	O
may	O
not	O
be	O
available	O
for	O
some	O
versions	O
of	O
Android	B-System
,	O
or	O
the	O
patches	O
were	O
never	O
installed	O
by	O
the	O
user	O
.	O

If	O
rooting	O
is	O
successful	O
,	O
the	O
attacker	O
has	O
full	O
control	O
of	O
the	O
device	O
and	O
can	O
execute	O
privileged	O
commands	O
remotely	O
.	O

After	O
achieving	O
root	O
access	O
,	O
Gooligan	B-Malware
downloads	O
a	O
new	O
,	O
malicious	O
module	O
from	O
the	O
C	O
&	O
C	O
server	O
and	O
installs	O
it	O
on	O
the	O
infected	O
device	O
.	O

This	O
module	O
injects	O
code	O
into	O
running	O
Google	B-System
Play	I-System
or	O
GMS	B-System
(	I-System
Google	I-System
Mobile	I-System
Services	I-System
)	I-System
to	O
mimic	O
user	O
behavior	O
so	O
Gooligan	B-Malware
can	O
avoid	O
detection	O
,	O
a	O
technique	O
first	O
seen	O
with	O
the	O
mobile	O
malware	O
HummingBad	B-Malware
.	O

The	O
module	O
allows	O
Gooligan	B-Malware
to	O
:	O
Steal	O
a	O
user	O
’	O
s	O
Google	B-Organization
email	O
account	O
and	O
authentication	O
token	O
information	O
Install	O
apps	O
from	O
Google	B-System
Play	I-System
and	O
rate	O
them	O
to	O
raise	O
their	O
reputation	O
Install	O
adware	O
to	O
generate	O
revenue	O
Ad	O
servers	O
,	O
which	O
don	O
’	O
t	O
know	O
whether	O
an	O
app	O
using	O
its	O
service	O
is	O
malicious	O
or	O
not	O
,	O
send	O
Gooligan	B-Malware
the	O
names	O
of	O
the	O
apps	O
to	O
download	O
from	O
Google	B-System
Play	I-System
.	O

After	O
an	O
app	O
is	O
installed	O
,	O
the	O
ad	O
service	O
pays	O
the	O
attacker	O
.	O

Then	O
the	O
malware	O
leaves	O
a	O
positive	O
review	O
and	O
a	O
high	O
rating	O
on	O
Google	B-System
Play	I-System
using	O
content	O
it	O
receives	O
from	O
the	O
C	O
&	O
C	O
server	O
.	O

Our	O
research	O
team	O
was	O
able	O
to	O
identify	O
several	O
instances	O
of	O
this	O
activity	O
by	O
cross-referencing	O
data	O
from	O
breached	O
devices	O
with	O
Google	B-System
Play	I-System
app	O
reviews	O
.	O

This	O
is	O
another	O
reminder	O
of	O
why	O
users	O
shouldn	O
’	O
t	O
rely	O
on	O
ratings	O
alone	O
to	O
decide	O
whether	O
to	O
trust	O
an	O
app	O
.	O

Similar	O
to	O
HummingBad	B-Malware
,	O
the	O
malware	O
also	O
fakes	O
device	O
identification	O
information	O
,	O
such	O
as	O
IMEI	O
and	O
IMSI	O
,	O
to	O
download	O
an	O
app	O
twice	O
while	O
seeming	O
like	O
the	O
installation	O
is	O
happening	O
on	O
a	O
different	O
device	O
,	O
thereby	O
doubling	O
the	O
potential	O
revenue	O
.	O

What	O
are	O
Google	B-Organization
authorization	O
tokens	O
?	O

A	O
Google	B-Organization
authorization	O
token	O
is	O
a	O
way	O
to	O
access	O
the	O
Google	B-Organization
account	O
and	O
the	O
related	O
services	O
of	O
a	O
user	O
.	O

It	O
is	O
issued	O
by	O
Google	B-Organization
once	O
a	O
user	O
successfully	O
logged	O
into	O
this	O
account	O
.	O

When	O
an	O
authorization	O
token	O
is	O
stolen	O
by	O
a	O
hacker	O
,	O
they	O
can	O
use	O
this	O
token	O
to	O
access	O
all	O
the	O
Google	B-Organization
services	O
related	O
to	O
the	O
user	O
,	O
including	O
Google	B-System
Play	I-System
,	O
Gmail	B-System
,	O
Google	B-System
Docs	I-System
,	O
Google	B-System
Drive	I-System
,	O
and	O
Google	B-System
Photos	I-System
.	O

While	O
Google	B-Organization
implemented	O
multiple	O
mechanisms	O
,	O
like	O
two-factor-authentication	O
,	O
to	O
prevent	O
hackers	O
from	O
compromising	O
Google	B-Organization
accounts	O
,	O
a	O
stolen	O
authorization	O
token	O
bypasses	O
this	O
mechanism	O
and	O
allows	O
hackers	O
the	O
desired	O
access	O
as	O
the	O
user	O
is	O
perceived	O
as	O
already	O
logged	O
in	O
.	O

Conclusion	O
Gooligan	B-Malware
has	O
breached	O
over	O
a	O
million	O
Google	B-Organization
accounts	O
.	O

We	O
believe	O
that	O
it	O
is	O
the	O
largest	O
Google	B-Malware
account	O
breach	O
to	O
date	O
,	O
and	O
we	O
are	O
working	O
with	O
Google	B-Organization
to	O
continue	O
the	O
investigation	O
.	O

We	O
encourage	O
Android	B-System
users	O
to	O
validate	O
whether	O
their	O
accounts	O
have	O
been	O
breached	O
.	O

Hacking	B-Organization
Team	I-Organization
Spying	O
Tool	O
Listens	O
to	O
Calls	O
By	O
:	O
Trend	B-Organization
Micro	I-Organization
July	O
21	O
,	O
2015	O
Following	O
news	O
that	O
iOS	B-System
devices	O
are	O
at	O
risk	O
of	O
spyware	O
related	O
to	O
the	O
Hacking	B-Organization
Team	I-Organization
,	O
the	O
saga	O
continues	O
into	O
the	O
Android	B-System
sphere	O
.	O

We	O
found	O
that	O
among	O
the	O
leaked	O
files	O
is	O
the	O
code	O
for	O
Hacking	O
Team	O
’	O
s	O
open-source	O
malware	O
suite	O
RCSAndroid	B-Malware
(	O
Remote	B-Malware
Control	I-Malware
System	I-Malware
Android	I-Malware
)	O
,	O
which	O
was	O
sold	O
by	O
the	O
company	O
as	O
a	O
tool	O
for	O
monitoring	O
targets	O
.	O

(	O
Researchers	O
have	O
been	O
aware	O
of	O
this	O
suite	O
as	O
early	O
as	O
2014	O
.	O

)	O
The	O
RCSAndroid	B-Malware
code	O
can	O
be	O
considered	O
one	O
of	O
the	O
most	O
professionally	O
developed	O
and	O
sophisticated	O
Android	B-System
malware	O
ever	O
exposed	O
.	O

The	O
leak	O
of	O
its	O
code	O
provides	O
cybercriminals	O
with	O
a	O
new	O
weaponized	O
resource	O
for	O
enhancing	O
their	O
surveillance	O
operations	O
.	O

Based	O
on	O
the	O
leaked	O
code	O
,	O
the	O
RCSAndroid	B-Malware
app	O
can	O
do	O
the	O
following	O
intrusive	O
routines	O
to	O
spy	O
on	O
targets	O
:	O
Capture	O
screenshots	O
using	O
the	O
“	O
screencap	O
”	O
command	O
and	O
framebuffer	O
direct	O
reading	O
Monitor	O
clipboard	O
content	O
Collect	O
passwords	O
for	O
Wi-Fi	O
networks	O
and	O
online	O
acco	O
;	O
.unts	O
,	O
including	O
Skype	B-System
,	O
Facebook	B-System
,	O
Twitter	B-System
,	O
Google	B-System
,	O
WhatsApp	B-System
,	O
Mail	B-System
,	O
and	O
LinkedIn	B-System
Record	O
using	O
the	O
microphone	O
Collect	O
SMS	O
,	O
MMS	O
,	O
and	O
Gmail	B-System
messages	O
Record	O
location	O
Gather	O
device	O
information	O
Capture	O
photos	O
using	O
the	O
front	O
and	O
back	O
cameras	O
Collect	O
contacts	O
and	O
decode	O

messages	O
from	O
IM	O
accounts	O
,	O
including	O
Facebook	B-System
Messenger	I-System
,	O
WhatsApp	B-System
,	O
Skype	B-System
,	O
Viber	B-System
,	O
Line	B-System
,	O
WeChat	B-System
,	O
Hangouts	B-System
,	O
Telegram	B-System
,	O
and	O
BlackBerry	B-System
Messenger	I-System
.	O

Capture	O
real-time	O
voice	O
calls	O
in	O
any	O
network	O
or	O
app	O
by	O
hooking	O
into	O
the	O
“	O
mediaserver	O
”	O
system	O
service	O
RCSAndroid	B-Malware
in	O
the	O
Wild	O
Our	O
analysis	O
reveals	O
that	O
this	O
RCSAndroid	B-Malware
(	O
AndroidOS_RCSAgent.HRX	B-Indicator
)	O
has	O
been	O
in	O
the	O
wild	O
since	O
2012	O
.	O

Traces	O
of	O
its	O
previous	O
uses	O
in	O
the	O
wild	O
were	O
found	O
inside	O
the	O
configuration	O
file	O
:	O
It	O
was	O
configured	O
to	O
use	O
a	O
Command-and-control	O
(	O
C	O
&	O
C	O
)	O
server	O
in	O
the	O
United	O
States	O
;	O
however	O
,	O
the	O
server	O
was	O
bought	O
from	O
a	O
host	O
service	O
provider	O
and	O
is	O
now	O
unavailable	O
.	O

It	O
was	O
configured	O
to	O
activate	O
via	O
SMS	O
sent	O
from	O
a	O
Czech	O
Republic	O
number	O
.	O

Attackers	O
can	O
send	O
SMS	O
with	O
certain	O
messages	O
to	O
activate	O
the	O
agent	O
and	O
trigger	O
corresponding	O
action	O
.	O

This	O
can	O
also	O
define	O
what	O
kind	O
of	O
evidences	O
to	O
collect	O
.	O

Based	O
on	O
emails	O
leaked	O
in	O
the	O
dump	O
,	O
a	O
number	O
of	O
Czech	O
firms	O
appear	O
to	O
be	O
in	O
business	O
with	O
the	O
Hacking	O
team	O
,	O
including	O
a	O
major	O
IT	O
partner	O
in	O
the	O
Olympic	O
Games	O
.	O

Dropping	O
Cluster	O
Bombs	O
RCSAndroid	B-Malware
is	O
a	O
threat	O
that	O
works	O
like	O
a	O
cluster	O
bomb	O
in	O
that	O
it	O
deploys	O
multiple	O
dangerous	O
exploits	O
and	O
uses	O
various	O
techniques	O
to	O
easily	O
infect	O
Android	B-System
devices	O
.	O

While	O
analyzing	O
the	O
code	O
,	O
we	O
found	O
that	O
the	O
whole	O
system	O
consists	O
of	O
four	O
critical	O
components	O
,	O
as	O
follows	O
:	O
penetration	O
solutions	O
,	O
ways	O
to	O
get	O
inside	O
the	O
device	O
,	O
either	O
via	O
SMS/email	O
or	O
a	O
legitimate	O
app	O
low-level	O
native	O
code	O
,	O
advanced	O
exploits	O
and	O
spy	O
tools	O
beyond	O
Android	B-System
’	O
s	O
security	O
framework	O
high-level	O
Java	O
agent	O
–	O
the	O
app	O
’	O
s	O
malicious	O
APK	O
command-and-control	O
(	O
C	O
&	O
C	O
)	O
servers	O
,	O
used	O
to	O
remotely	O
send/receive	O
malicious	O
commands	O
Attackers	O
use	O
two	O
methods	O
to	O
get	O
targets	O
to	O
download	O
RCSAndroid	B-Malware
.	O

The	O
first	O
method	O
is	O
to	O
send	O
a	O
specially	O
crafted	O
URL	O
to	O
the	O
target	O
via	O
SMS	O
or	O
email	O
.	O

The	O
URL	O
will	O
trigger	O
exploits	O
for	O
arbitrary	B-Vulnerability
memory	I-Vulnerability
read	I-Vulnerability
(	I-Vulnerability
CVE-2012-2825	I-Vulnerability
)	I-Vulnerability
and	O
heap	B-Vulnerability
buffer	I-Vulnerability
overflow	I-Vulnerability
(	I-Vulnerability
CVE-2012-2871	I-Vulnerability
)	I-Vulnerability
vulnerabilities	O
in	O
the	O
default	O
browsers	O
of	O
Android	B-System
versions	I-System
4.0	I-System
Ice	I-System
Cream	I-System
Sandwich	I-System
to	O
4.3	B-System
Jelly	I-System
Bean	I-System
,	O
allowing	O
another	O
local	O
privilege	O
escalation	O
exploit	O
to	O
execute	O
.	O

When	O
root	O
privilege	O
is	O
gained	O
,	O
a	O
shell	O
backdoor	O
and	O
malicious	O
RCSAndroid	B-Malware
agent	O
APK	O
file	O
will	O
be	O
installed	O
The	O
second	O
method	O
is	O
to	O
use	O
a	O
stealthy	O
backdoor	O
app	O
such	O
as	O
ANDROIDOS_HTBENEWS.A	B-Malware
,	O
which	O
was	O
designed	O
to	O
bypass	O
Google	B-System
Play	I-System
.	O

The	O
role	O
of	O
ANDROIDOS_HTBENEWS.A	B-Malware
and	O
the	O
malicious	O
APK	O
mentioned	O
in	O
the	O
first	O
method	O
is	O
to	O
exploit	O
a	O
local	B-Vulnerability
privilege	I-Vulnerability
escalation	I-Vulnerability
vulnerability	I-Vulnerability
in	O
Android	O
devices	O
.	O

Hacking	O
Team	O
has	O
been	O
known	O
to	O
use	O
both	O
CVE-2014-3153	B-Vulnerability
and	O
CVE-2013-6282	B-Vulnerability
in	O
their	O
attacks	O
.	O

The	O
said	O
exploits	O
will	O
root	O
the	O
device	O
and	O
install	O
a	O
shell	O
backdoor	O
.	O

The	O
shell	O
backdoor	O
then	O
installs	O
the	O
RCSAndroid	B-Malware
agent	O
.	O

This	O
agent	O
has	O
two	O
core	O
modules	O
,	O
the	O
Evidence	O
Collector	O
and	O
the	O
Event	O
Action	O
Trigger	O
.	O

The	O
Evidence	O
Collector	O
module	O
is	O
responsible	O
for	O
the	O
spying	O
routines	O
outlined	O
above	O
.	O

One	O
of	O
its	O
most	O
notable	O
routines	O
is	O
capturing	O
voice	O
calls	O
in	O
real	O
time	O
by	O
hooking	O
into	O
the	O
“	O
mediaserver	O
”	O
system	O
service	O
.	O

The	O
basic	O
idea	O
is	O
to	O
hook	O
the	O
voice	O
call	O
process	O
in	O
mediaserver	O
.	O

Take	O
voice	O
call	O
playback	O
process	O
for	O
example	O
.	O

The	O
mediaserver	O
will	O
first	O
builds	O
a	O
new	O
unique	O
track	O
,	O
start	O
to	O
play	O
the	O
track	O
,	O
loop	O
play	O
all	O
audio	O
buffer	O
,	O
then	O
finally	O
stop	O
the	O
playback	O
.	O

The	O
raw	O
wave	O
audio	O
buffer	O
frame	O
can	O
be	O
dumped	O
in	O
the	O
getNextBuffer	O
(	O
)	O
function	O
.	O

With	O
the	O
help	O
of	O
the	O
open-source	O
Android	B-System
Dynamic	O
Binary	O
Instrumentation	O
Toolkit	O
and	O
root	O
privilege	O
,	O
it	O
is	O
possible	O
to	O
intercept	O
any	O
function	O
execution	O
.	O

The	O
Event	O
Action	O
Trigger	O
module	O
triggers	O
malicious	O
actions	O
based	O
on	O
certain	O
events	O
.	O

These	O
events	O
can	O
be	O
based	O
on	O
time	O
,	O
charging	O
or	O
battery	O
status	O
,	O
location	O
,	O
connectivity	O
,	O
running	O
apps	O
,	O
focused	O
app	O
,	O
SIM	O
card	O
status	O
,	O
SMS	O
received	O
with	O
keywords	O
,	O
and	O
screen	O
turning	O
on	O
.	O

According	O
to	O
the	O
configuration	O
pattern	O
,	O
these	O
actions	O
are	O
registered	O
to	O
certain	O
events	O
:	O
Sync	O
configuration	O
data	O
,	O
upgrade	O
modules	O
,	O
and	O
download	O
new	O
payload	O
(	O
This	O
uses	O
transport	O
protocol	O
ZProtocol	O
encrypted	O
by	O
AES/CBC/PKCS5Padding	O
algorithm	O
to	O
communicate	O
with	O
the	O
C	O
&	O
C	O
server	O
.	O

)	O
Upload	O
and	O
purge	O
collected	O
evidence	O
Destroy	O
device	O
by	O
resetting	O
locking	O
password	O
Execute	O
shell	O
commands	O
Send	O
SMS	O
with	O
defined	O
content	O
or	O
location	O
Disable	O
network	O
Disable	O
root	O
Uninstall	O
bot	O
To	O
avoid	O
detection	O
and	O
removal	O
of	O
the	O
agent	O
app	O
in	O
the	O
device	O
memory	O
,	O
the	O
RCSAndroid	B-Malware
suite	O
also	O
detects	O
emulators	O
or	O
sandboxes	O
,	O
obfuscates	O
code	O
using	O
DexGuard	B-System
,	O
uses	O
ELF	O
string	O
obfuscator	O
,	O
and	O
adjusts	O
the	O
OOM	O
(	O
out-of-memory	O
)	O
value	O
.	O

Interestingly	O
,	O
one	O
unused	O
feature	O
of	O
the	O
app	O
is	O
its	O
ability	O
to	O
manipulate	O
data	O
in	O
the	O
Android	B-System
package	O
manager	O
to	O
add	O
and	O
remove	O
permissions	O
and	O
components	O
as	O
well	O
as	O
hide	O
the	O
app	O
icon	O
.	O

Recommendations	O
Popular	O
mobile	O
platforms	O
like	O
Android	B-System
are	O
common	O
targets	O
for	O
organized	O
or	O
commercialized	O
monitoring	O
operations	O
.	O

Attackers	O
know	O
that	O
rooting	O
devices	O
via	O
malware	O
exploits	O
is	O
an	O
effective	O
means	O
to	O
control	O
devices	O
and	O
gather	O
information	O
from	O
them	O
.	O

In	O
a	O
root	O
broken	O
device	O
,	O
security	O
is	O
a	O
fairy	O
tale	O
.	O

Take	O
note	O
of	O
the	O
following	O
best	O
practices	O
to	O
prevent	O
this	O
threat	O
from	O
getting	O
in	O
your	O
device	O
:	O
Disable	O
app	O
installations	O
from	O
unknown	O
,	O
third-party	O
sources	O
.	O

Constantly	O
update	O
your	O
Android	B-System
devices	O
to	O
the	O
latest	O
version	O
to	O
help	O
prevent	O
exploits	O
,	O
especially	O
in	O
the	O
case	O
of	O
RCSAndroid	B-Malware
which	O
can	O
affect	O
only	O
up	O
to	O
version	O
4.4.4	B-System
KitKat	I-System
.	O

Note	O
,	O
however	O
,	O
that	O
based	O
on	O
the	O
leak	O
mail	O
from	O
a	O
customer	O
inquiry	O
,	O
Hacking	B-Organization
Team	I-Organization
was	O
in	O
the	O
process	O
of	O
developing	O
exploits	O
for	O
Android	B-System
5.0	I-System
Lollipop	I-System
.	O

Install	O
a	O
mobile	O
security	O
solution	O
to	O
secure	O
your	O
device	O
from	O
threats	O
.	O

The	O
leaked	O
RCSAndroid	B-Malware
code	I-Malware
is	O
a	O
commercial	O
weapon	O
now	O
in	O
the	O
wild	O
.	O

Mobile	O
users	O
are	O
called	O
on	O
to	O
be	O
on	O
top	O
of	O
this	O
news	O
and	O
be	O
on	O
guard	O
for	O
signs	O
of	O
monitoring	O
.	O

Some	O
indicators	O
may	O
come	O
in	O
the	O
form	O
of	O
peculiar	O
behavior	O
such	O
as	O
unexpected	O
rebooting	O
,	O
finding	O
unfamiliar	O
apps	O
installed	O
,	O
or	O
instant	O
messaging	O
apps	O
suddenly	O
freezing	O
.	O

Should	O
a	O
device	O
become	O
infected	O
,	O
this	O
backdoor	O
can	O
not	O
be	O
removed	O
without	O
root	O
privilege	O
.	O

Users	O
may	O
be	O
required	O
the	O
help	O
of	O
their	O
device	O
manufacturer	O
to	O
get	O
support	O
for	O
firmware	O
flashing	O
.	O

Trend	B-Organization
Micro	I-Organization
offers	O
security	O
for	O
Android	B-System
mobile	O
devices	O
through	O
Mobile	B-System
Security	I-System
for	I-System
Android™	I-System
to	O
protect	O
against	O
these	O
types	O
of	O
attacks	O
.	O

Find	O
out	O
more	O
about	O
the	O
7	O
Android	B-System
Security	O
Hacks	O
You	O
Need	O
to	O
Do	O
Right	O
Now	O
to	O
keep	O
your	O
mobile	O
data	O
safe	O
.	O

Update	O
as	O
of	O
July	O
23	O
,	O
2015	O
1:00	O
AM	O
PDT	O
(	O
UTC-7	O
)	O
We	O
have	O
added	O
a	O
link	O
to	O
a	O
previous	O
report	O
discussing	O
this	O
threat	O
.	O

Timeline	O
of	O
posts	O
related	O
to	O
the	O
Hacking	O
Team	O
DATE	O
UPDATE	O
July	O
5	O
The	O
Italian	O
company	O
Hacking	B-Organization
Team	I-Organization
was	O
hacked	O
,	O
with	O
more	O
than	O
400GB	O
of	O
confidential	O
company	O
data	O
made	O
available	O
to	O
the	O
public	O
.	O

July	O
7	O
Three	O
exploits	O
–	O
two	O
for	O
Flash	B-System
Player	I-System
and	O
one	O
for	O
the	O
Windows	B-System
kernel—were	O
initially	O
found	O
in	O
the	O
information	O
dump	O
.	O

One	O
of	O
these	O
[	O
CVE-2015-5119	B-Vulnerability
]	O
was	O
a	O
Flash	O
zero-day	O
.	O

The	O
Windows	B-Vulnerability
kernel	I-Vulnerability
vulnerability	I-Vulnerability
(	O
CVE-2015-2387	B-Vulnerability
)	O
existed	O
in	O
the	O
open	O
type	O
font	O
manager	O
module	O
(	O
ATMFD.dll	B-Indicator
)	O
and	O
can	O
be	O
exploited	O
to	O
bypass	O
the	O
sandbox	O
mitigation	O
mechanism	O
.	O

The	O
Flash	B-System
zero-day	O
exploit	O
(	O
CVE-2015-5119	B-Vulnerability
)	O
was	O
added	O
into	O
the	O
Angler	B-Malware
Exploit	I-Malware
Kit	I-Malware
and	O
Nuclear	B-Malware
Exploit	I-Malware
Pack	I-Malware
.	O

It	O
was	O
also	O
used	O
in	O
limited	O
attacks	O
in	O
Korea	O
and	O
Japan	O
.	O

July	O
11	O
Two	O
new	O
Flash	B-Vulnerability
zero-day	I-Vulnerability
vulnerabilities	I-Vulnerability
,	O
CVE-2015-5122	B-Vulnerability
and	O
CVE-2015-5123	B-Vulnerability
,	O
were	O
found	O
in	O
the	O
hacking	O
team	O
dump	O
.	O

July	O
13	O
Further	O
analysis	O
of	O
the	O
hacking	O
team	O
dump	O
revealed	O
that	O
the	O
company	O
used	O
UEFI	B-Malware
BIOS	I-Malware
rootkit	I-Malware
to	O
keep	O
their	O
Remote	B-Malware
Control	I-Malware
System	I-Malware
(	I-Malware
RCS	I-Malware
)	I-Malware
agent	O
installed	O
in	O
their	O
targets	O
’	O
systems	O
.	O

July	O
14	O
A	O
new	O
zero-day	B-Vulnerability
vulnerability	I-Vulnerability
(	O
CVE-2015-2425	B-Vulnerability
)	O
was	O
found	O
in	O
Internet	B-System
Explorer	I-System
.	O

July	O
16	O
On	O
the	O
mobile	O
front	O
,	O
a	O
fake	O
news	O
app	O
designed	O
to	O
bypass	O
Google	B-System
Play	I-System
was	O
discovered	O
.	O

July	O
20	O
A	O
new	O
zero-day	B-Vulnerability
vulnerability	I-Vulnerability
(	O
CVE-2015-2426	B-Vulnerability
)	O
was	O
found	O
in	O
Windows	B-System
,	O
which	O
Microsoft	B-Organization
fixed	O
in	O
an	O
out-of-band	O
patch	O
.	O

July	O
21	O
Analysis	O
of	O
the	O
RCSAndroid	B-Malware
spying	O
tool	O
revealed	O
that	O
Hacking	O
Team	O
can	O
listen	O
to	O
calls	O
and	O
roots	O
devices	O
to	O
get	O
in	O
.	O

July	O
28	O
A	O
recent	O
campaign	O
compromised	O
Taiwan	O
and	O
Hong	O
Kong	O
sites	O
to	O
deliver	O
Flash	B-System
exploits	O
related	O
to	O
Hacking	B-Organization
Team	I-Organization
.	O

Android	B-System
users	O
warned	O
of	O
malware	O
attack	O
spreading	O
via	O
SMS	O
FEB	O
16	O
,	O
2016	O
Security	O
researchers	O
are	O
warning	O
owners	O
of	O
Android	B-System
smartphones	O
about	O
a	O
new	O
malware	O
attack	O
,	O
spreading	O
via	O
SMS	O
text	O
messages	O
.	O

As	O
the	O
team	O
at	O
Scandinavian	O
security	O
group	O
CSIS	B-Organization
describes	O
,	O
malware	O
known	O
as	O
MazarBOT	B-Malware
is	O
being	O
distributed	O
via	O
SMS	O
in	O
Denmark	O
and	O
is	O
likely	O
to	O
also	O
be	O
encountered	O
in	O
other	O
countries	O
.	O

Victims	O
’	O
first	O
encounter	O
with	O
the	O
malware	O
reportedly	O
comes	O
via	O
an	O
unsolicited	O
text	O
message	O
that	O
their	O
Android	B-System
smartphone	I-System
receives	O
.	O

The	O
txt	O
message	O
uses	O
social	O
engineering	O
to	O
dupe	O
unsuspecting	O
users	O
into	O
clicking	O
on	O
a	O
link	O
to	O
a	O
downloadable	O
Android	B-System
application	O
.	O

CSIS	B-Organization
provided	O
a	O
(	O
sanitised	O
)	O
version	O
of	O
a	O
typical	O
message	O
to	O
warn	O
users	O
what	O
to	O
look	O
out	O
for	O
:	O
“	O
You	O
have	O
received	O
a	O
multimedia	O
message	O
from	O
+	O
[	O
country	O
code	O
]	O
[	O
sender	O
number	O
]	O
Follow	O
the	O
link	O
http	B-Indicator
:	I-Indicator
//www.mmsforyou	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
net/mms.apk	I-Indicator
to	O
view	O
the	O
message	O
”	O
Once	O
the	O
APK	O
package	O
is	O
downloaded	O
,	O
potential	O
victims	O
are	O
urged	O
to	O
grant	O
the	O
malicious	O
app	O
a	O
wide	O
range	O
of	O
permissions	O
on	O
their	O
Android	O
device	O
:	O
App	O
permissions	O
SEND_SMS	O
RECEIVE_BOOT_COMPLETED	O
INTERNET	O
SYSTEM_ALERT_WINDOW	O
WRITE_SMS	O
ACCESS_NETWORK_STATE	O
WAKE_LOCK	O
GET_TASKS	O
CALL_PHONE	O
RECEIVE_SMS	O
READ_PHONE_STATE	O
READ_SMS	O
ERASE_PHONE	O
Once	O
installed	O
,	O
MazarBOT	B-Malware
downloads	O
a	O
copy	O
of	O

Tor	B-System
onto	O
users	O
’	O
Android	B-System
smartphones	O
and	O
uses	O
it	O
to	O
connect	O
anonymously	O
to	O
the	O
net	O
before	O
sending	O
a	O
text	O
message	O
containing	O
the	O
victim	O
’	O
s	O
location	O
to	O
an	O
Iranian	O
mobile	O
phone	O
number	O
.	O

With	O
the	O
malware	O
now	O
in	O
place	O
,	O
a	O
number	O
of	O
actions	O
can	O
be	O
performed	O
,	O
including	O
allowing	O
attackers	O
to	O
secretly	O
monitor	O
and	O
control	O
smartphones	O
via	O
a	O
backdoor	O
,	O
send	O
messages	O
to	O
premium-rate	O
numbers	O
,	O
and	O
intercept	O
two-factor	O
authentication	O
codes	O
sent	O
by	O
online	O
banking	O
apps	O
and	O
the	O
like	O
.	O

In	O
fact	O
,	O
with	O
full	O
access	O
to	O
the	O
compromised	O
Android	B-System
smartphone	I-System
,	O
the	O
opportunities	O
for	O
criminals	O
to	O
wreak	O
havoc	O
are	O
significant	O
–	O
such	O
as	O
erasing	O
infected	O
phones	O
or	O
launching	O
man-in-the-middle	O
(	O
MITM	O
)	O
attacks	O
.	O

In	O
its	O
analysis	O
,	O
CSIS	B-Organization
notes	O
that	O
MazarBOT	B-Malware
was	O
reported	O
by	O
Recorded	B-Organization
Future	I-Organization
last	O
November	O
as	O
being	O
actively	O
sold	O
in	O
Russian	O
underground	O
forums	O
and	O
intriguingly	O
,	O
the	O
malware	O
will	O
not	O
activate	O
on	O
Android	B-System
devices	O
configured	O
with	O
Russian	O
language	O
settings	O
.	O

This	O
,	O
in	O
itself	O
,	O
does	O
not	O
prove	O
that	O
the	O
perpetrators	O
of	O
the	O
malware	O
campaign	O
are	O
based	O
in	O
Russia	O
,	O
but	O
it	O
certainly	O
sounds	O
as	O
if	O
that	O
is	O
a	O
strong	O
possibility	O
.	O

Malware	O
authors	O
in	O
the	O
past	O
have	O
often	O
coded	O
a	O
“	O
safety	O
net	O
”	O
into	O
their	O
malware	O
to	O
prevent	O
them	O
from	O
accidentally	O
infecting	O
their	O
own	O
computers	O
and	O
devices	O
.	O

For	O
more	O
detailed	O
information	O
about	O
the	O
threat	O
,	O
check	O
out	O
the	O
blog	O
post	O
from	O
CSIS	B-Organization
.	O

And	O
,	O
of	O
course	O
,	O
remember	O
to	O
always	O
be	O
wary	O
of	O
unsolicited	O
,	O
unusual	O
text	O
messages	O
and	O
installing	O
apps	O
from	O
third-party	O
sources	O
on	O
your	O
Android	B-System
smartphone	I-System
.	O

Coronavirus	B-System
Update	I-System
App	I-System
Leads	O
to	O
Project	B-Malware
Spy	I-Malware
Android	B-System
and	O
iOS	B-System
Spyware	O
We	O
discovered	O
a	O
cyberespionage	O
campaign	O
we	O
have	O
named	O
Project	B-Malware
Spy	I-Malware
infecting	O
Android	B-System
and	O
iOS	B-System
devices	O
with	O
spyware	O
by	O
using	O
the	O
coronavirus	O
disease	O
(	O
Covid-19	O
)	O
as	O
a	O
lure	O
.	O

By	O
:	O
Tony	O
Bao	O
,	O
Junzhi	O
Lu	O
April	O
14	O
,	O
2020	O
We	O
discovered	O
a	O
potential	O
cyberespionage	O
campaign	O
,	O
which	O
we	O
have	O
named	O
Project	B-Malware
Spy	I-Malware
,	O
that	O
infects	O
Android	B-System
and	O
iOS	B-System
devices	O
with	O
spyware	O
(	O
detected	O
by	O
Trend	B-Organization
Micro	I-Organization
as	O
AndroidOS_ProjectSpy.HRX	B-Indicator
and	O
IOS_ProjectSpy.A	B-Indicator
,	O
respectively	O
)	O
.	O

Project	B-Malware
Spy	I-Malware
uses	O
the	O
ongoing	O
coronavirus	O
pandemic	O
as	O
a	O
lure	O
,	O
posing	O
as	O
an	O
app	O
called	O
Coronavirus	O
Updates	O
.	O

We	O
also	O
found	O
similarities	O
in	O
two	O
older	O
samples	O
disguised	O
as	O
a	O
Google	B-Organization
service	O
and	O
,	O
subsequently	O
,	O
as	O
a	O
music	O
app	O
after	O
further	O
investigation	O
.	O

However	O
,	O
we	O
have	O
noted	O
a	O
significantly	O
small	O
number	O
of	O
downloads	O
of	O
the	O
app	O
in	O
Pakistan	O
,	O
India	O
,	O
Afghanistan	O
,	O
Bangladesh	O
,	O
Iran	O
,	O
Saudi	O
Arabia	O
,	O
Austria	O
,	O
Romania	O
,	O
Grenada	O
,	O
and	O
Russia	O
.	O

Project	B-Malware
Spy	I-Malware
routine	O
At	O
the	O
end	O
of	O
March	O
2020	O
,	O
we	O
came	O
across	O
an	O
app	O
masquerading	O
as	O
a	O
coronavirus	O
update	O
app	O
,	O
which	O
we	O
named	O
Project	B-Malware
Spy	I-Malware
based	O
on	O
the	O
login	O
page	O
of	O
its	O
backend	O
server	O
.	O

This	O
app	O
carries	O
a	O
number	O
of	O
the	O
capabilities	O
:	O
Upload	O
GSM	B-System
,	O
WhatsApp	B-System
,	O
Telegram	B-System
,	O
Facebook	B-System
,	O
and	O
Threema	B-System
messages	O
Upload	O
voice	O
notes	O
,	O
contacts	O
stored	O
,	O
accounts	O
,	O
call	O
logs	O
,	O
location	O
information	O
,	O
and	O
images	O
Upload	O
the	O
expanded	O
list	O
of	O
collected	O
device	O
information	O
(	O
e.g.	O
,	O
IMEI	O
,	O
product	O
,	O
board	O
,	O
manufacturer	O
,	O
tag	O
,	O
host	O
,	O
Android	B-System
version	O
,	O
application	O
version	O
,	O
name	O
,	O
model	O
brand	O
,	O
user	O
,	O
serial	O
,	O
hardware	O
,	O
bootloader	O
,	O
and	O
device	O
ID	O
)	O
Upload	O
SIM	O
information	O
(	O
e.g.	O

,	O
IMSI	O
,	O
operator	O
code	O
,	O
country	O
,	O
MCC-mobile	O
country	O
,	O
SIM	O
serial	O
,	O
operator	O
name	O
,	O
and	O
mobile	O
number	O
)	O
Upload	O
wifi	O
information	O
(	O
e.g.	O
,	O
SSID	O
,	O
wifi	O
speed	O
,	O
and	O
MAC	O
address	O
)	O
Upload	O
other	O
information	O
(	O
e.g.	O
,	O
display	O
,	O
date	O
,	O
time	O
,	O
fingerprint	O
,	O
created	O
at	O
,	O
and	O
updated	O
at	O
)	O
The	O
app	O
is	O
capable	O
of	O
stealing	O
messages	O
from	O
popular	O
messaging	O
apps	O
by	O
abusing	O
the	O
notification	O
permissions	O
to	O
read	O
the	O
notification	O
content	O
and	O
saving	O
it	O
to	O
the	O
database	O
.	O

It	O
requests	O
permission	O
to	O
access	O
the	O
additional	O
storage	O
.	O

Project	B-Malware
Spy	I-Malware
’	O
s	O
earlier	O
versions	O
Searching	O
for	O
the	O
domain	O
in	O
our	O
sample	O
database	O
,	O
we	O
found	O
that	O
the	O
coronavirus	O
update	O
app	O
appears	O
to	O
be	O
the	O
latest	O
version	O
of	O
another	O
sample	O
that	O
we	O
detected	O
in	O
May	O
2019	O
.	O

The	O
first	O
version	O
of	O
Project	B-Malware
Spy	I-Malware
(	O
detected	O
by	O
Trend	B-Organization
Micro	I-Organization
as	O
AndroidOS_SpyAgent.HRXB	B-Indicator
)	O
had	O
the	O
following	O
capabilities	O
:	O
Collect	O
device	O
and	O
system	O
information	O
(	O
i.e.	O
,	O
IMEI	O
,	O
device	O
ID	O
,	O
manufacturer	O
,	O
model	O
and	O
phone	O
number	O
)	O
,	O
location	O
information	O
,	O
contacts	O
stored	O
,	O
and	O
call	O
logs	O
Collect	O
and	O
send	O
SMS	O
Take	O
pictures	O
via	O
the	O
camera	O
Upload	O
recorded	O
MP4	O
files	O
Monitor	O
calls	O
Searching	O
further	O
,	O
we	O
also	O
found	O
another	O
sample	O
that	O
could	O
be	O
the	O
second	O
version	O
of	O
Project	O
Spy	O
.	O

This	O
version	O
appeared	O
as	O
Wabi	O
Music	O
,	O
and	O
copied	O
a	O
popular	O
video-sharing	O
social	O
networking	O
service	O
as	O
its	O
backend	O
login	O
page	O
.	O

In	O
this	O
second	O
version	O
,	O
the	O
developer	O
’	O
s	O
name	O
listed	O
was	O
“	O
concipit1248	O
”	O
in	O
Google	B-System
Play	I-System
,	O
and	O
may	O
have	O
been	O
active	O
between	O
May	O
2019	O
to	O
February	O
2020	O
.	O

This	O
app	O
appears	O
to	O
have	O
become	O
unavailable	O
on	O
Google	B-System
Play	I-System
in	O
March	O
2020	O
.	O

The	O
second	O
Project	B-Malware
Spy	I-Malware
version	O
has	O
similar	O
capabilities	O
to	O
the	O
first	O
version	O
,	O
with	O
the	O
addition	O
of	O
the	O
following	O
:	O
Stealing	O
notification	O
messages	O
sent	O
from	O
WhatsApp	B-System
,	O
Facebook	B-System
,	O
and	O
Telegram	B-System
Abandoning	O
the	O
FTP	O
mode	O
of	O
uploading	O
the	O
recorded	O
images	O
Aside	O
from	O
changing	O
the	O
app	O
’	O
s	O
supposed	O
function	O
and	O
look	O
,	O
the	O
second	O
and	O
third	O
versions	O
’	O
codes	O
had	O
little	O
differences	O
.	O

Potentially	O
malicious	O
iOS	B-System
connection	O
Using	O
the	O
codes	O
and	O
“	O
Concipit1248	O
”	O
to	O
check	O
for	O
more	O
versions	O
,	O
we	O
found	O
two	O
other	O
apps	O
in	O
the	O
App	B-System
Store	I-System
.	O

Further	O
analysis	O
of	O
the	O
iOS	O
app	O
“	O
Concipit1248	O
”	O
showed	O
that	O
the	O
server	O
used	O
,	O
spy	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
cashnow	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ee	I-Indicator
,	O
is	O
the	O
same	O
one	O
used	O
in	O
the	O
Android	B-System
version	O
of	O
Project	B-System
Spy	I-System
.	O

However	O
,	O
although	O
the	O
“	O
Concipit1248	O
”	O
app	O
requested	O
permissions	O
to	O
open	O
the	O
device	O
camera	O
and	O
read	O
photos	O
,	O
the	O
code	O
only	O
can	O
upload	O
a	O
self-contained	O
PNG	O
file	O
to	O
a	O
remote	O
sever	O
.	O

This	O
may	O
imply	O
the	O
“	O
Concipit1248	O
”	O
app	O
is	O
still	O
incubating	O
.	O

The	O
other	O
iOS	B-System
app	O
“	O
Concipit	O
Shop	O
”	O
from	O
the	O
same	O
developer	O
appeared	O
normal	O
and	O
was	O
last	O
updated	O
on	O
November	O
2019	O
.	O

Apple	B-Organization
has	O
confirmed	O
that	O
the	O
iOS	B-System
apps	O
are	O
not	O
functioning	O
based	O
on	O
analysis	O
of	O
the	O
codes	O
,	O
and	O
stated	O
that	O
the	O
sandbox	O
is	O
able	O
to	O
detect	O
and	O
block	O
these	O
malicious	O
behaviors	O
.	O

Conclusion	O
The	O
“	O
Corona	O
Updates	O
”	O
app	O
had	O
relatively	O
low	O
downloads	O
in	O
Pakistan	O
,	O
India	O
,	O
Afghanistan	O
,	O
Bangladesh	O
,	O
Iran	O
,	O
Saudi	O
Arabia	O
,	O
Austria	O
,	O
Romania	O
,	O
Grenada	O
,	O
and	O
Russia	O
.	O

Perhaps	O
the	O
app	O
’	O
s	O
false	O
capabilities	O
also	O
fueled	O
the	O
low	O
number	O
of	O
downloads	O
.	O

It	O
also	O
appears	O
the	O
apps	O
may	O
still	O
be	O
in	O
development	O
or	O
incubation	O
,	O
maybe	O
waiting	O
for	O
a	O
“	O
right	O
time	O
”	O
to	O
inject	O
the	O
malicious	O
codes	O
.	O

It	O
’	O
s	O
also	O
possible	O
that	O
the	O
apps	O
are	O
being	O
used	O
to	O
test	O
other	O
possible	O
techniques	O
.	O

A	O
possible	O
indication	O
for	O
timing	O
might	O
be	O
when	O
the	O
app	O
reaches	O
a	O
specific	O
number	O
of	O
downloads	O
or	O
infected	O
devices	O
.	O

The	O
coding	O
style	O
suggests	O
that	O
the	O
cybercriminals	O
behind	O
this	O
campaign	O
are	O
amateurs	O
.	O

The	O
incomplete	O
iOS	B-System
codes	O
used	O
in	O
this	O
campaign	O
may	O
have	O
been	O
bought	O
while	O
other	O
capabilities	O
appear	O
to	O
have	O
been	O
added	O
.	O

This	O
may	O
also	O
explain	O
the	O
timing	O
in	O
between	O
the	O
apps	O
becoming	O
fully	O
functional	O
and	O
“	O
incubation.	O
”	O
As	O
this	O
is	O
a	O
group	O
we	O
have	O
not	O
observed	O
before	O
,	O
we	O
will	O
continue	O
monitoring	O
this	O
campaign	O
for	O
further	O
developments	O
.	O

Users	O
are	O
cautioned	O
to	O
research	O
and	O
check	O
reviews	O
before	O
they	O
download	O
apps	O
.	O

Observe	O
and	O
look	O
at	O
the	O
app	O
’	O
s	O
display	O
and	O
text	O
,	O
stated	O
functions	O
,	O
reviews	O
from	O
other	O
users	O
,	O
and	O
requested	O
permissions	O
before	O
downloading	O
.	O

Make	O
sure	O
that	O
all	O
other	O
apps	O
installed	O
and	O
the	O
device	O
operating	O
systems	O
are	O
updated	O
to	O
the	O
latest	O
version	O
.	O

Indicators	O
of	O
Compromise	O
(	O
IoCs	O
)	O
SHA256	O
Detection	O
e394e53e53cd9047d6cff184ac333ef7698a34b777ae3aac82c2c669ef661dfe	B-Indicator
AndroidOS_SpyAgent.HRXB	B-Indicator
e8d4713e43241ab09d40c2ae8814302f77de76650ccf3e7db83b3ac8ad41f9fa	B-Indicator
AndroidOS_ProjectSpy.HRX	O
29b0d86ae68d83f9578c3f36041df943195bc55a7f3f1d45a9c23f145d75af9d	B-Indicator

AndroidOS_ProjectSpy.HRX	O
3a15e7b8f4e35e006329811a6a2bf291d449884a120332f24c7e3ca58d0fbbdb	B-Indicator
IOS_ProjectSpy.A	B-Indicator
URLs	O
cashnow	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ee	I-Indicator
Backend	O
server	B-Indicator
ftp	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
XXXX	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
Backend	O
server	B-Indicator
spy	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
cashnow	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ee	I-Indicator
Backend	O
server	B-Indicator
xyz	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
cashnow	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ee	I-Indicator
Backend	O
server	O
October	O
8	O
,	O
2020	O
Sophisticated	O
new	O
Android	B-System
malware	O
marks	O
the	O
latest	O
evolution	O
of	O
mobile	O
ransomware	O
Attackers	O
are	O
persistent	O
and	O
motivated	O
to	O
continuously	O
evolve	O
–	O
and	O
no	O
platform	O
is	O
immune	O
.	O

That	O
is	O
why	O
Microsoft	B-Organization
has	O
been	O
working	O
to	O
extend	O
its	O
industry-leading	O
endpoint	O
protection	O
capabilities	O
beyond	O
Windows	B-System
.	O

The	O
addition	O
of	O
mobile	O
threat	O
defense	O
into	O
these	O
capabilities	O
means	O
that	O
Microsoft	B-System
Defender	I-System
for	O
Endpoint	O
(	O
previously	O
Microsoft	B-System
Defender	I-System
Advanced	I-System
Threat	I-System
Protection	I-System
)	O
now	O
delivers	O
protection	O
on	O
all	O
major	O
platforms	O
.	O

Microsoft	B-Organization
’	O
s	O
mobile	O
threat	O
defense	O
capabilities	O
further	O
enrich	O
the	O
visibility	O
that	O
organizations	O
have	O
on	O
threats	O
in	O
their	O
networks	O
,	O
as	O
well	O
as	O
provide	O
more	O
tools	O
to	O
detect	O
and	O
respond	O
to	O
threats	O
across	O
domains	O
and	O
across	O
platforms	O
.	O

Like	O
all	O
of	O
Microsoft	B-Organization
’	O
s	O
security	O
solutions	O
,	O
these	O
new	O
capabilities	O
are	O
likewise	O
backed	O
by	O
a	O
global	O
network	O
of	O
threat	O
researchers	O
and	O
security	O
experts	O
whose	O
deep	O
understanding	O
of	O
the	O
threat	O
landscape	O
guide	O
the	O
continuous	O
innovation	O
of	O
security	O
features	O
and	O
ensure	O
that	O
customers	O
are	O
protected	O
from	O
ever-evolving	O
threats	O
.	O

For	O
example	O
,	O
we	O
found	O
a	O
piece	O
of	O
a	O
particularly	O
sophisticated	O
Android	B-System
ransomware	O
with	O
novel	O
techniques	O
and	O
behavior	O
,	O
exemplifying	O
the	O
rapid	O
evolution	O
of	O
mobile	O
threats	O
that	O
we	O
have	O
also	O
observed	O
on	O
other	O
platforms	O
.	O

The	O
mobile	O
ransomware	O
,	O
detected	O
by	O
Microsoft	B-System
Defender	I-System
for	O
Endpoint	O
as	O
AndroidOS/MalLocker.B	B-Indicator
,	O
is	O
the	O
latest	O
variant	O
of	O
a	O
ransomware	O
family	O
that	O
’	O
s	O
been	O
in	O
the	O
wild	O
for	O
a	O
while	O
but	O
has	O
been	O
evolving	O
non-stop	O
.	O

This	O
ransomware	O
family	O
is	O
known	O
for	O
being	O
hosted	O
on	O
arbitrary	O
websites	O
and	O
circulated	O
on	O
online	O
forums	O
using	O
various	O
social	O
engineering	O
lures	O
,	O
including	O
masquerading	O
as	O
popular	O
apps	O
,	O
cracked	O
games	O
,	O
or	O
video	O
players	O
.	O

The	O
new	O
variant	O
caught	O
our	O
attention	O
because	O
it	O
’	O
s	O
an	O
advanced	O
malware	O
with	O
unmistakable	O
malicious	O
characteristic	O
and	O
behavior	O
and	O
yet	O
manages	O
to	O
evade	O
many	O
available	O
protections	O
,	O
registering	O
a	O
low	O
detection	O
rate	O
against	O
security	O
solutions	O
.	O

As	O
with	O
most	O
Android	B-System
ransomware	O
,	O
this	O
new	O
threat	O
doesn	O
’	O
t	O
actually	O
block	O
access	O
to	O
files	O
by	O
encrypting	O
them	O
.	O

Instead	O
,	O
it	O
blocks	O
access	O
to	O
devices	O
by	O
displaying	O
a	O
screen	O
that	O
appears	O
over	O
every	O
other	O
window	O
,	O
such	O
that	O
the	O
user	O
can	O
’	O
t	O
do	O
anything	O
else	O
.	O

The	O
said	O
screen	O
is	O
the	O
ransom	O
note	O
,	O
which	O
contains	O
threats	O
and	O
instructions	O
to	O
pay	O
the	O
ransom	O
.	O

What	O
’	O
s	O
innovative	O
about	O
this	O
ransomware	O
is	O
how	O
it	O
displays	O
its	O
ransom	O
note	O
.	O

In	O
this	O
blog	O
,	O
we	O
’	O
ll	O
detail	O
the	O
innovative	O
ways	O
in	O
which	O
this	O
ransomware	O
surfaces	O
its	O
ransom	O
note	O
using	O
Android	B-System
features	O
we	O
haven	O
’	O
t	O
seen	O
leveraged	O
by	O
malware	O
before	O
,	O
as	O
well	O
as	O
incorporating	O
an	O
open-source	O
machine	O
learning	O
module	O
designed	O
for	O
context-aware	O
cropping	O
of	O
its	O
ransom	O
note	O
.	O

New	O
scheme	O
,	O
same	O
goal	O
In	O
the	O
past	O
,	O
Android	B-System
ransomware	O
used	O
a	O
special	O
permission	O
called	O
“	O
SYSTEM_ALERT_WINDOW	O
”	O
to	O
display	O
their	O
ransom	O
note	O
.	O

Apps	O
that	O
have	O
this	O
permission	O
can	O
draw	O
a	O
window	O
that	O
belongs	O
to	O
the	O
system	O
group	O
and	O
can	O
’	O
t	O
be	O
dismissed	O
.	O

No	O
matter	O
what	O
button	O
is	O
pressed	O
,	O
the	O
window	O
stays	O
on	O
top	O
of	O
all	O
other	O
windows	O
.	O

The	O
notification	O
was	O
intended	O
to	O
be	O
used	O
for	O
system	O
alerts	O
or	O
errors	O
,	O
but	O
Android	B-System
threats	O
misused	O
it	O
to	O
force	O
the	O
attacker-controlled	O
UI	O
to	O
fully	O
occupy	O
the	O
screen	O
,	O
blocking	O
access	O
to	O
the	O
device	O
.	O

Attackers	O
create	O
this	O
scenario	O
to	O
persuade	O
users	O
to	O
pay	O
the	O
ransom	O
so	O
they	O
can	O
gain	O
back	O
access	O
to	O
the	O
device	O
.	O

To	O
catch	O
these	O
threats	O
,	O
security	O
solutions	O
used	O
heuristics	O
that	O
focused	O
on	O
detecting	O
this	O
behavior	O
.	O

Google	B-Organization
later	O
implemented	O
platform-level	O
changes	O
that	O
practically	O
eliminated	O
this	O
attack	O
surface	O
.	O

These	O
changes	O
include	O
:	O
Removing	O
the	O
SYSTEM_ALERT_WINDOW	O
error	O
and	O
alert	O
window	O
types	O
,	O
and	O
introducing	O
a	O
few	O
other	O
types	O
as	O
replacement	O
Elevating	O
the	O
permission	O
status	O
of	O
SYSTEM_ALERT_WINDOW	O
to	O
special	O
permission	O
by	O
putting	O
it	O
into	O
the	O
“	O
above	O
dangerous	O
”	O
category	O
,	O
which	O
means	O
that	O
users	O
have	O
to	O
go	O
through	O
many	O
screens	O
to	O
approve	O
apps	O
that	O
ask	O
for	O
permission	O
,	O
instead	O
of	O
just	O
one	O
click	O
Introducing	O
an	O
overlay	O
kill	O
switch	O
on	O
Android	B-System
8.0	I-System
and	O
later	O
that	O
users	O
can	O
activate	O
anytime	O
to	O
deactivate	O
a	O
system	O
alert	O
window	O
To	O
adapt	O
,	O
Android	B-System
malware	O
evolved	O
to	O
misusing	O

other	O
features	O
,	O
but	O
these	O
aren	O
’	O
t	O
as	O
effective	O
.	O

For	O
example	O
,	O
some	O
strains	O
of	O
ransomware	O
abuse	O
accessibility	O
features	O
,	O
a	O
method	O
that	O
could	O
easily	O
alarm	O
users	O
because	O
accessibility	O
is	O
a	O
special	O
permission	O
that	O
requires	O
users	O
to	O
go	O
through	O
several	O
screens	O
and	O
accept	O
a	O
warning	O
that	O
the	O
app	O
will	O
be	O
able	O
to	O
monitor	O
activity	O
via	O
accessibility	O
services	O
.	O

Other	O
ransomware	O
families	O
use	O
infinite	O
loops	O
of	O
drawing	O
non-system	O
windows	B-System
,	O
but	O
in	O
between	O
drawing	O
and	O
redrawing	O
,	O
it	O
’	O
s	O
possible	O
for	O
users	O
to	O
go	O
to	O
settings	O
and	O
uninstall	O
the	O
offending	O
app	O
.	O

The	O
new	O
Android	B-Malware
ransomware	O
variant	O
overcomes	O
these	O
barriers	O
by	O
evolving	O
further	O
than	O
any	O
Android	B-Malware
malware	O
we	O
’	O
ve	O
seen	O
before	O
.	O

To	O
surface	O
its	O
ransom	O
note	O
,	O
it	O
uses	O
a	O
series	O
of	O
techniques	O
that	O
take	O
advantage	O
of	O
the	O
following	O
components	O
on	O
Android	B-System
:	O
The	O
“	O
call	O
”	O
notification	O
,	O
among	O
several	O
categories	O
of	O
notifications	O
that	O
Android	B-System
supports	O
,	O
which	O
requires	O
immediate	O
user	O
attention	O
.	O

The	O
“	O
onUserLeaveHint	O
(	O
)	O
”	O
callback	O
method	O
of	O
the	O
Android	B-System
Activity	I-System
(	O
i.e.	O
,	O
the	O
typical	O
GUI	O
screen	O
the	O
user	O
sees	O
)	O
is	O
called	O
as	O
part	O
of	O
the	O
activity	O
lifecycle	O
when	O
the	O
activity	O
is	O
about	O
to	O
go	O
into	O
the	O
background	O
as	O
a	O
result	O
of	O
user	O
choice	O
,	O
for	O
example	O
,	O
when	O
the	O
user	O
presses	O
the	O
Home	O
key	O
.	O

The	O
malware	O
connects	O
the	O
dots	O
and	O
uses	O
these	O
two	O
components	O
to	O
create	O
a	O
special	O
type	O
of	O
notification	O
that	O
triggers	O
the	O
ransom	O
screen	O
via	O
the	O
callback	O
.	O

As	O
the	O
code	O
snippet	O
shows	O
,	O
the	O
malware	O
creates	O
a	O
notification	O
builder	O
and	O
then	O
does	O
the	O
following	O
:	O
setCategory	O
(	O
“	O
call	O
”	O
)	O
–	O
This	O
means	O
that	O
the	O
notification	O
is	O
built	O
as	O
a	O
very	O
important	O
notification	O
that	O
needs	O
special	O
privilege	O
.	O

setFullScreenIntent	O
(	O
)	O
–	O
This	O
API	O
wires	O
the	O
notification	O
to	O
a	O
GUI	O
so	O
that	O
it	O
pops	O
up	O
when	O
the	O
user	O
taps	O
on	O
it	O
.	O

At	O
this	O
stage	O
,	O
half	O
the	O
job	O
is	O
done	O
for	O
the	O
malware	O
.	O

However	O
,	O
the	O
malware	O
wouldn	O
’	O
t	O
want	O
to	O
depend	O
on	O
user	O
interaction	O
to	O
trigger	O
the	O
ransomware	O
screen	O
,	O
so	O
,	O
it	O
adds	O
another	O
functionality	O
of	O
Android	B-System
callback	O
:	O
As	O
the	O
code	O
snippet	O
shows	O
,	O
the	O
malware	O
overrides	O
the	O
onUserLeaveHint	O
(	O
)	O
callback	O
function	O
of	O
Activity	O
class	O
.	O

The	O
function	O
onUserLeaveHint	O
(	O
)	O
is	O
called	O
whenever	O
the	O
malware	O
screen	O
is	O
pushed	O
to	O
background	O
,	O
causing	O
the	O
in-call	O
Activity	O
to	O
be	O
automatically	O
brought	O
to	O
the	O
foreground	O
.	O

Recall	O
that	O
the	O
malware	O
hooked	O
the	O
RansomActivity	O
intent	O
with	O
the	O
notification	O
that	O
was	O
created	O
as	O
a	O
“	O
call	O
”	O
type	O
notification	O
.	O

This	O
creates	O
a	O
chain	O
of	O
events	O
that	O
triggers	O
the	O
automatic	O
pop-up	O
of	O
the	O
ransomware	O
screen	O
without	O
doing	O
infinite	O
redraw	O
or	O
posing	O
as	O
system	O
window	O
.	O

Machine	O
learning	O
module	O
indicates	O
continuous	O
evolution	O
As	O
mentioned	O
,	O
this	O
ransomware	O
is	O
the	O
latest	O
variant	O
of	O
a	O
malware	O
family	O
that	O
has	O
undergone	O
several	O
stages	O
of	O
evolution	O
.	O

The	O
knowledge	O
graph	O
below	O
shows	O
the	O
various	O
techniques	O
this	O
ransomware	O
family	O
has	O
been	O
seen	O
using	O
,	O
including	O
abusing	O
the	O
system	O
alert	O
window	O
,	O
abusing	O
accessibility	O
features	O
,	O
and	O
,	O
more	O
recently	O
,	O
abusing	O
notification	O
services	O
.	O

This	O
ransomware	O
family	O
’	O
s	O
long	O
history	O
tells	O
us	O
that	O
its	O
evolution	O
is	O
far	O
from	O
over	O
.	O

We	O
expect	O
it	O
to	O
churn	O
out	O
new	O
variants	O
with	O
even	O
more	O
sophisticated	O
techniques	O
.	O

In	O
fact	O
,	O
recent	O
variants	O
contain	O
code	O
forked	O
from	O
an	O
open-source	O
machine	O
learning	O
module	O
used	O
by	O
developers	O
to	O
automatically	O
resize	O
and	O
crop	O
images	O
based	O
on	O
screen	O
size	O
,	O
a	O
valuable	O
function	O
given	O
the	O
variety	O
of	O
Android	B-System
devices	O
.	O

The	O
frozen	O
TinyML	B-System
model	O
is	O
useful	O
for	O
making	O
sure	O
images	O
fit	O
the	O
screen	O
without	O
distortion	O
.	O

In	O
the	O
case	O
of	O
this	O
ransomware	O
,	O
using	O
the	O
model	O
would	O
ensure	O
that	O
its	O
ransom	O
note—typically	O
fake	O
police	O
notice	O
or	O
explicit	O
images	O
supposedly	O
found	O
on	O
the	O
device—would	O
appear	O
less	O
contrived	O
and	O
more	O
believable	O
,	O
increasing	O
the	O
chances	O
of	O
the	O
user	O
paying	O
for	O
the	O
ransom	O
.	O

The	O
library	O
that	O
uses	O
tinyML	B-System
is	O
not	O
yet	O
wired	O
to	O
the	O
malware	O
’	O
s	O
functionalities	O
,	O
but	O
its	O
presence	O
in	O
the	O
malware	O
code	O
indicates	O
the	O
intention	O
to	O
do	O
so	O
in	O
future	O
variants	O
.	O

We	O
will	O
continue	O
to	O
monitor	O
this	O
ransomware	O
family	O
to	O
ensure	O
customers	O
are	O
protected	O
and	O
to	O
share	O
our	O
findings	O
and	O
insights	O
to	O
the	O
community	O
for	O
broad	O
protection	O
against	O
these	O
evolving	O
mobile	O
threats	O
.	O

Protecting	O
organizations	O
from	O
threats	O
across	O
domains	O
and	O
platforms	O
Mobile	O
threats	O
continue	O
to	O
rapidly	O
evolve	O
,	O
with	O
attackers	O
continuously	O
attempting	O
to	O
sidestep	O
technological	O
barriers	O
and	O
creatively	O
find	O
ways	O
to	O
accomplish	O
their	O
goal	O
,	O
whether	O
financial	O
gain	O
or	O
finding	O
an	O
entry	O
point	O
to	O
broader	O
network	O
compromise	O
.	O

This	O
new	O
mobile	O
ransomware	O
variant	O
is	O
an	O
important	O
discovery	O
because	O
the	O
malware	O
exhibits	O
behaviors	O
that	O
have	O
not	O
been	O
seen	O
before	O
and	O
could	O
open	O
doors	O
for	O
other	O
malware	O
to	O
follow	O
.	O

It	O
reinforces	O
the	O
need	O
for	O
comprehensive	O
defense	O
powered	O
by	O
broad	O
visibility	O
into	O
attack	O
surfaces	O
as	O
well	O
as	O
domain	O
experts	O
who	O
track	O
the	O
threat	O
landscape	O
and	O
uncover	O
notable	O
threats	O
that	O
might	O
be	O
hiding	O
amidst	O
massive	O
threat	O
data	O
and	O
signals	O
.	O

Microsoft	B-System
Defender	I-System
for	O
Endpoint	O
on	O
Android	B-System
,	O
now	O
generally	O
available	O
,	O
extends	O
Microsoft	B-Organization
’	O
s	O
industry-leading	O
endpoint	O
protection	O
to	O
Android	B-System
.	O

It	O
detects	O
this	O
ransomware	O
(	O
AndroidOS/MalLocker.B	B-Indicator
)	O
,	O
as	O
well	O
as	O
other	O
malicious	O
apps	O
and	O
files	O
using	O
cloud-based	O
protection	O
powered	O
by	O
deep	O
learning	O
and	O
heuristics	O
,	O
in	O
addition	O
to	O
content-based	O
detection	O
.	O

It	O
also	O
protects	O
users	O
and	O
organizations	O
from	O
other	O
mobile	O
threats	O
,	O
such	O
as	O
mobile	O
phishing	O
,	O
unsafe	O
network	O
connections	O
,	O
and	O
unauthorized	O
access	O
to	O
sensitive	O
data	O
.	O

Learn	O
more	O
about	O
our	O
mobile	O
threat	O
defense	O
capabilities	O
in	O
Microsoft	B-System
Defender	I-System
for	O
Endpoint	O
on	O
Android	B-System
.	O

Malware	O
,	O
phishing	O
,	O
and	O
other	O
threats	O
detected	O
by	O
Microsoft	B-System
Defender	I-System
for	O
Endpoint	O
are	O
reported	O
to	O
the	O
Microsoft	B-Organization
Defender	I-Organization
Security	I-Organization
Center	I-Organization
,	O
allowing	O
SecOps	O
to	O
investigate	O
mobile	O
threats	O
along	O
with	O
endpoint	O
signals	O
from	O
Windows	B-System
and	O
other	O
platforms	O
using	O
Microsoft	B-System
Defender	I-System
for	O
Endpoint	O
’	O
s	O
rich	O
set	O
of	O
tools	O
for	O
detection	O
,	O
investigation	O
,	O
and	O
response	O
.	O

Threat	O
data	O
from	O
endpoints	O
are	O
combined	O
with	O
signals	O
from	O
email	O
and	O
data	O
,	O
identities	O
,	O
and	O
apps	O
in	O
Microsoft	B-System
365	I-System
Defender	I-System
(	O
previously	O
Microsoft	B-System
Threat	I-System
Protection	I-System
)	O
,	O
which	O
orchestrates	O
detection	O
,	O
prevention	O
,	O
investigation	O
,	O
and	O
response	O
across	O
domains	O
,	O
providing	O
coordinated	O
defense	O
.	O

Microsoft	B-System
Defender	I-System
for	O
Endpoint	O
on	O
Android	B-System
further	O
enriches	O
organizations	O
’	O
visibility	O
into	O
malicious	O
activity	O
,	O
empowering	O
them	O
to	O
comprehensively	O
prevent	O
,	O
detect	O
,	O
and	O
respond	O
to	O
against	O
attack	O
sprawl	O
and	O
cross-domain	O
incidents	O
.	O

Technical	O
analysis	O
Obfuscation	O
On	O
top	O
of	O
recreating	O
ransomware	O
behavior	O
in	O
ways	O
we	O
haven	O
’	O
t	O
seen	O
before	O
,	O
the	O
Android	B-System
malware	O
variant	O
uses	O
a	O
new	O
obfuscation	O
technique	O
unique	O
to	O
the	O
Android	B-System
platform	O
.	O

One	O
of	O
the	O
tell-tale	O
signs	O
of	O
an	O
obfuscated	O
malware	O
is	O
the	O
absence	O
of	O
code	O
that	O
defines	O
the	O
classes	O
declared	O
in	O
the	O
manifest	O
file	O
.	O

The	O
classes.dex	O
has	O
implementation	O
for	O
only	O
two	O
classes	O
:	O
The	O
main	O
application	O
class	O
gCHotRrgEruDv	O
,	O
which	O
is	O
involved	O
when	O
the	O
application	O
opens	O
A	O
helper	O
class	O
that	O
has	O
definition	O
for	O
custom	O
encryption	O
and	O
decryption	O
This	O
means	O
that	O
there	O
’	O
s	O
no	O
code	O
corresponding	O
to	O
the	O
services	O
declared	O
in	O
the	O
manifest	O
file	O
:	O
Main	O
Activity	O
,	O
Broadcast	O
Receivers	O
,	O
and	O
Background	O
.	O

How	O
does	O
the	O
malware	O
work	O
without	O
code	O
for	O
these	O
key	O
components	O
?	O

As	O
is	O
characteristic	O
for	O
obfuscated	O
threats	O
,	O
the	O
malware	O
has	O
encrypted	O
binary	O
code	O
stored	O
in	O
the	O
Assets	O
folder	O
:	O
When	O
the	O
malware	O
runs	O
for	O
the	O
first	O
time	O
,	O
the	O
static	O
block	O
of	O
the	O
main	O
class	O
is	O
run	O
.	O

The	O
code	O
is	O
heavily	O
obfuscated	O
and	O
made	O
unreadable	O
through	O
name	O
mangling	O
and	O
use	O
of	O
meaningless	O
variable	O
names	O
:	O
Decryption	O
with	O
a	O
twist	O
The	O
malware	O
uses	O
an	O
interesting	O
decryption	O
routine	O
:	O
the	O
string	O
values	O
passed	O
to	O
the	O
decryption	O
function	O
do	O
not	O
correspond	O
to	O
the	O
decrypted	O
value	O
,	O
they	O
correspond	O
to	O
junk	O
code	O
to	O
simply	O
hinder	O
analysis	O
.	O

On	O
Android	B-System
,	O
an	O
Intent	O
is	O
a	O
software	O
mechanism	O
that	O
allows	O
users	O
to	O
coordinate	O
the	O
functions	O
of	O
different	O
Activities	O
to	O
achieve	O
a	O
task	O
.	O

It	O
’	O
s	O
a	O
messaging	O
object	O
that	O
can	O
be	O
used	O
to	O
request	O
an	O
action	O
from	O
another	O
app	O
component	O
.	O

The	O
Intent	O
object	O
carries	O
a	O
string	O
value	O
as	O
“	O
action	O
”	O
parameter	O
.	O

The	O
malware	O
creates	O
an	O
Intent	O
inside	O
the	O
decryption	O
function	O
using	O
the	O
string	O
value	O
passed	O
as	O
the	O
name	O
for	O
the	O
Intent	O
.	O

It	O
then	O
decrypts	O
a	O
hardcoded	O
encrypted	O
value	O
and	O
sets	O
the	O
“	O
action	O
”	O
parameter	O
of	O
the	O
Intent	O
using	O
the	O
setAction	O
API	O
.	O

Once	O
this	O
Intent	O
object	O
is	O
generated	O
with	O
the	O
action	O
value	O
pointing	O
to	O
the	O
decrypted	O
content	O
,	O
the	O
decryption	O
function	O
returns	O
the	O
Intent	O
object	O
to	O
the	O
callee	O
.	O

The	O
callee	O
then	O
invokes	O
the	O
getAction	O
method	O
to	O
get	O
the	O
decrypted	O
content	O
.	O

Payload	O
deployment	O
Once	O
the	O
static	O
block	O
execution	O
is	O
complete	O
,	O
the	O
Android	B-System
Lifecycle	I-System
callback	O
transfers	O
the	O
control	O
to	O
the	O
OnCreate	O
method	O
of	O
the	O
main	O
class	O
.	O

Malware	O
code	O
showing	O
onCreate	O
method	O
Figure	O
9.	O
onCreate	O
method	O
of	O
the	O
main	O
class	O
decrypting	O
the	O
payload	O
Next	O
,	O
the	O
malware-defined	O
function	O
decryptAssetToDex	O
(	O
a	O
meaningful	O
name	O
we	O
assigned	O
during	O
analysis	O
)	O
receives	O
the	O
string	O
“	O
CuffGmrQRT	B-Indicator
”	O
as	O
the	O
first	O
argument	O
,	O
which	O
is	O
the	O
name	O
of	O
the	O
encrypted	O
file	O
stored	O
in	O
the	O
Assets	O
folder	O
.	O

Malware	O
code	O
showing	O
decryption	O
of	O
assets	O
Figure	O
10	O
.	O

Decrypting	O
the	O
assets	O
After	O
being	O
decrypted	O
,	O
the	O
asset	O
turns	O
into	O
the	O
.dex	O
file	O
.	O

This	O
is	O
a	O
notable	O
behavior	O
that	O
is	O
characteristic	O
of	O
this	O
ransomware	O
family	O
.	O

Comparison	O
of	O
code	O
of	O
Asset	O
file	O
before	O
and	O
after	O
decryption	O
Figure	O
11	O
.	O

Asset	O
file	O
before	O
and	O
after	O
decryption	O
Once	O
the	O
encrypted	O
executable	O
is	O
decrypted	O
and	O
dropped	O
in	O
the	O
storage	O
,	O
the	O
malware	O
has	O
the	O
definitions	O
for	O
all	O
the	O
components	O
it	O
declared	O
in	O
the	O
manifest	O
file	O
.	O

It	O
then	O
starts	O
the	O
final	O
detonator	O
function	O
to	O
load	O
the	O
dropped	O
.dex	O
file	O
into	O
memory	O
and	O
triggers	O
the	O
main	O
payload	O
.	O

Malware	O
code	O
showing	O
loading	O
of	O
decrypted	O
dex	O
file	O
Figure	O
12	O
.	O

Loading	O
the	O
decrypted	O
.dex	O
file	O
into	O
memory	O
and	O
triggering	O
the	O
main	O
payload	O
Main	O
payload	O
When	O
the	O
main	O
payload	O
is	O
loaded	O
into	O
memory	O
,	O
the	O
initial	O
detonator	O
hands	O
over	O
the	O
control	O
to	O
the	O
main	O
payload	O
by	O
invoking	O
the	O
method	O
XoqF	O
(	O
which	O
we	O
renamed	O
to	O
triggerInfection	O
during	O
analysis	O
)	O
from	O
the	O
gvmthHtyN	O
class	O
(	O
renamed	O
to	O
PayloadEntry	O
)	O
.	O

Malware	O
code	O
showing	O
handover	O
from	O
initial	O
module	O
to	O
main	O
payload	O
Figure	O
13	O
.	O

Handover	O
from	O
initial	O
module	O
to	O
the	O
main	O
payload	O
As	O
mentioned	O
,	O
the	O
initial	O
handover	O
component	O
called	O
triggerInfection	O
with	O
an	O
instance	O
of	O
appObj	O
and	O
a	O
method	O
that	O
returns	O
the	O
value	O
for	O
the	O
variable	O
config	O
.	O

Malware	O
code	O
showing	O
definition	O
of	O
populateConfigMap	O
Figure	O
14	O
.	O

Definition	O
of	O
populateConfigMap	O
,	O
which	O
loads	O
the	O
map	O
with	O
values	O
Correlating	O
the	O
last	O
two	O
steps	O
,	O
one	O
can	O
observe	O
that	O
the	O
malware	O
payload	O
receives	O
the	O
configuration	O
for	O
the	O
following	O
properties	O
:	O
number	O
–	O
The	O
default	O
number	O
to	O
be	O
send	O
to	O
the	O
server	O
(	O
in	O
case	O
the	O
number	O
is	O
not	O
available	O
from	O
the	O
device	O
)	O
api	O
–	O
The	O
API	O
key	O
url	O
–	O
The	O
URL	O
to	O
be	O
used	O
in	O
WebView	O
to	O
display	O
on	O
the	O
ransom	O
note	O
The	O
malware	O
saves	O
this	O
configuration	O
to	O
the	O
shared	O
preferences	O
of	O
the	O
app	O
data	O
and	O
then	O
it	O
sets	O
up	O
all	O
the	O
Broadcast	O
Receivers	O
.	O

This	O
action	O
registers	O
code	O
components	O
to	O
get	O
notified	O
when	O
certain	O
system	O
events	O
happen	O
.	O

This	O
is	O
done	O
in	O
the	O
function	O
initComponents	O
.	O

Malware	O
code	O
showing	O
initializing	O
broadcast	O
receiver	O
Figure	O
15	O
.	O

Initializing	O
the	O
BroadcastReceiver	O
against	O
system	O
events	O
From	O
this	O
point	O
on	O
,	O
the	O
malware	O
execution	O
is	O
driven	O
by	O
callback	O
functions	O
that	O
are	O
triggered	O
on	O
system	O
events	O
like	O
connectivity	O
change	O
,	O
unlocking	O
the	O
phone	O
,	O
elapsed	O
time	O
interval	O
,	O
and	O
others	O
.	O

Lookout	B-Organization
researchers	O
have	O
identified	O
a	O
new	O
,	O
highly	O
targeted	O
surveillanceware	O
family	O
known	O
as	O
Desert	B-Malware
Scorpion	I-Malware
in	O
the	O
Google	B-System
Play	I-System
Store	I-System
.	O

Lookout	B-Organization
notified	O
Google	B-Organization
of	O
the	O
finding	O
and	O
Google	B-Organization
removed	O
the	O
app	O
immediately	O
while	O
also	O
taking	O
action	O
on	O
it	O
in	O
Google	B-System
Play	I-System
Protect	I-System
.	O

The	O
app	O
ties	O
together	O
two	O
malware	O
families	O
-	O
Desert	B-Malware
Scorpion	I-Malware
and	O
another	O
targeted	O
surveillanceware	O
family	O
named	O
FrozenCell	B-Malware
-	O
that	O
we	O
believe	O
are	O
being	O
developed	O
by	O
a	O
single	O
,	O
evolving	O
surveillanceware	O
actor	O
called	O
APT-C-23	B-Malware
targeting	O
individuals	O
in	O
the	O
Middle	O
East	O
.	O

We	O
've	O
seen	O
this	O
actor	O
rely	O
heavily	O
on	O
phishing	O
campaigns	O
to	O
trick	O
victims	O
into	O
downloading	O
their	O
malicious	O
apps	O
,	O
specifically	O
on	O
Facebook	B-System
.	O

Even	O
sophisticated	O
actors	O
are	O
using	O
lower	O
cost	O
,	O
less	O
technologically	O
impressive	O
means	O
like	O
phishing	O
to	O
spread	O
their	O
malware	O
because	O
it	O
's	O
cheap	O
and	O
very	O
effective	O
,	O
especially	O
on	O
mobile	O
devices	O
where	O
there	O
are	O
more	O
ways	O
to	O
interact	O
with	O
a	O
victim	O
(	O
messaging	O
apps	O
,	O
social	O
media	O
apps	O
,	O
etc	O
.	O

)	O
,	O
and	O
less	O
screen	O
real	O
estate	O
for	O
victims	O
to	O
identify	O
potential	O
indicators	O
of	O
a	O
threat	O
.	O

Lookout	B-Organization
customers	O
are	O
protected	O
against	O
this	O
threat	O
and	O
additionally	O
we	O
have	O
included	O
a	O
list	O
of	O
IOCs	O
at	O
the	O
end	O
of	O
this	O
report	O
.	O

The	O
potential	O
actor	O
and	O
who	O
they	O
target	O
Our	O
current	O
analysis	O
strongly	O
suggests	O
Desert	B-Malware
Scorpion	I-Malware
is	O
being	O
deployed	O
in	O
targeted	O
attacks	O
against	O
Middle	O
Eastern	O
individuals	O
of	O
interest	O
specifically	O
those	O
in	O
Palestine	O
and	O
has	O
also	O
been	O
highlighted	O
by	O
other	O
researchers	O
.	O

We	O
have	O
been	O
able	O
to	O
tie	O
the	O
malware	O
to	O
a	O
long-running	O
Facebook	B-Organization
profile	O
that	O
we	O
observed	O
promoting	O
the	O
first	O
stage	O
of	O
this	O
family	O
,	O
a	O
malicious	O
chat	O
application	O
called	O
Dardesh	B-Malware
via	O
links	O
to	O
Google	B-System
Play	I-System
.	O

The	O
Lookout	B-Organization
Threat	I-Organization
Intelligence	I-Organization
team	O
identified	O
that	O
this	O
same	O
Facebook	B-Organization
profile	O
has	O
also	O
posted	O
Google	B-System
Drive	I-System
links	O
to	O
Android	B-System
malware	O
belonging	O
to	O
the	O
FrozenCell	B-Malware
family	O
attributed	O
to	O
APT-C-27	B-Indicator
.	O

These	O
factors	O
,	O
in	O
combination	O
with	O
the	O
fact	O
that	O
the	O
command	O
and	O
control	O
infrastructure	O
used	O
by	O
Frozen	B-Malware
Cell	I-Malware
and	O
Desert	B-Malware
Scorpion	I-Malware
resides	O
in	O
similar	O
IP	O
blocks	O
,	O
supports	O
the	O
theory	O
that	O
the	O
same	O
actor	O
is	O
responsible	O
for	O
operating	O
,	O
if	O
not	O
developing	O
,	O
both	O
families	O
.	O

What	O
it	O
does	O
The	O
surveillance	O
functionality	O
of	O
Desert	B-Malware
Scorpion	I-Malware
resides	O
in	O
a	O
second	O
stage	O
payload	O
that	O
can	O
only	O
be	O
downloaded	O
if	O
the	O
victim	O
has	O
downloaded	O
,	O
installed	O
,	O
and	O
interacted	O
with	O
the	O
first-stage	O
chat	O
application	O
.	O

The	O
chat	O
application	O
acts	O
as	O
a	O
dropper	O
for	O
this	O
second-stage	O
payload	O
app	O
.	O

At	O
the	O
time	O
of	O
writing	O
Lookout	B-Organization
has	O
observed	O
two	O
updates	O
to	O
the	O
Dardesh	B-Malware
application	O
,	O
the	O
first	O
on	O
February	O
26	O
and	O
the	O
second	O
on	O
March	O
28	O
.	O

The	O
malicious	O
capabilities	O
observed	O
in	O
the	O
second	O
stage	O
include	O
the	O
following	O
:	O
Upload	O
attacker-specified	O
files	O
to	O
C2	O
servers	O
Get	O
list	O
of	O
installed	O
applications	O
Get	O
device	O
metadata	O
Inspect	O
itself	O
to	O
get	O
a	O
list	O
of	O
launchable	O
activities	O
Retrieves	O
PDF	O
,	O
txt	O
,	O
doc	O
,	O
xls	O
,	O
xlsx	O
,	O
ppt	O
,	O
pptx	O
files	O
found	O
on	O
external	O
storage	O
Send	O
SMS	O
Retrieve	O
text	O
messages	O
Track	O
device	O
location	O
Handle	O
limited	O
attacker	O
commands	O
via	O
out	O
of	O
band	O
text	O
messages	O
Record	O
surrounding	O
audio	O
Record	O
calls	O
Record	O
video	O
Retrieve	O
account	O
information	O
such	O
as	O
email	O
addresses	O
Retrieve	O
contacts	O
Removes	O
copies	O
of	O
itself	O
if	O

any	O
additional	O
APKs	O
are	O
downloaded	O
to	O
external	O
storage	O
.	O

Call	O
an	O
attacker-specified	O
number	O
Uninstall	O
apps	O
Check	O
if	O
a	O
device	O
is	O
rooted	O
Hide	O
its	O
icon	O
Retrieve	O
list	O
of	O
files	O
on	O
external	O
storage	O
If	O
running	O
on	O
a	O
Huawei	O
device	O
it	O
will	O
attempt	O
to	O
add	O
itself	O
to	O
the	O
protected	O
list	O
of	O
apps	O
able	O
to	O
run	O
with	O
the	O
screen	O
off	O
Encrypts	O
some	O
exfiltrated	O
data	O
Desert	B-Malware
Scorpion	I-Malware
's	O
second	O
stage	O
masquerades	O
as	O
a	O
generic	O
"	O
settings	O
''	O
application	O
.	O

Curiously	O
,	O
several	O
of	O
these	O
have	O
included	O
the	O
world	O
"	O
Fateh	O
''	O
in	O
their	O
package	O
name	O
,	O
which	O
may	O
be	O
referring	O
to	O
the	O
Fatah	B-Organization
political	O
party	O
.	O

Such	O
references	O
would	O
be	O
in	O
line	O
with	O
FrozenCell	B-Malware
's	O
phishing	O
tactics	O
in	O
which	O
they	O
used	O
file	O
names	O
to	O
lure	O
people	O
associated	O
with	O
the	O
political	O
party	O
to	O
open	O
malicious	O
documents	O
.	O

Desert	B-Malware
Scorpion	I-Malware
's	O
second	O
stage	O
is	O
capable	O
of	O
installing	O
another	O
non-malicious	O
application	O
(	O
included	O
in	O
the	O
second	O
stage	O
)	O
which	O
is	O
highly	O
specific	O
to	O
the	O
Fatah	B-Organization
political	O
party	O
and	O
supports	O
the	O
targeting	O
theory	O
.	O

The	O
Lookout	B-Organization
Threat	I-Organization
Intelligence	I-Organization
team	O
is	O
increasingly	O
seeing	O
the	O
same	O
tradecraft	O
,	O
tactics	O
,	O
and	O
procedures	O
that	O
APT-C-23	B-Malware
favors	O
being	O
used	O
by	O
other	O
actors	O
.	O

The	O
approach	O
of	O
separating	O
malicious	O
functionality	O
out	O
into	O
separate	O
stages	O
that	O
are	O
later	O
downloaded	O
during	O
execution	O
and	O
not	O
present	O
in	O
the	O
initial	O
app	O
published	O
to	O
the	O
Google	B-System
Play	I-System
Store	I-System
,	O
combined	O
with	O
social	O
engineering	O
delivered	O
via	O
social	O
media	O
platforms	O
like	O
Facebook	B-Organization
,	O
requires	O
minimal	O
investment	O
in	O
comparison	O
to	O
premium	O
tooling	O
like	O
Pegasus	B-Malware
or	O
FinFisher	B-Malware
.	O

As	O
we	O
've	O
seen	O
with	O
actors	O
like	O
Dark	B-Malware
Caracal	I-Malware
,	O
this	O
low	O
cost	O
,	O
low	O
sophistication	O
approach	O
that	O
relies	O
heavily	O
upon	O
social	O
engineering	O
has	O
still	O
been	O
shown	O
to	O
be	O
highly	O
successful	O
for	O
those	O
operating	O
such	O
campaigns	O
.	O

Given	O
previous	O
operational	O
security	O
errors	O
from	O
this	O
actor	O
in	O
the	O
past	O
which	O
resulted	O
in	O
exfiltrated	O
content	O
being	O
publicly	O
accessible	O
Lookout	B-Organization
Threat	I-Organization
Intelligence	I-Organization
is	O
continuing	O
to	O
map	O
out	O
infrastructure	O
and	O
closely	O
monitor	O
their	O
continued	O
evolution	O
.	O

Virulent	B-Malware
Android	B-System
malware	O
returns	O
,	O
gets	O
>	O
2	O
million	O
downloads	O
on	O
Google	B-System
Play	I-System
HummingWhale	B-Malware
is	O
back	O
with	O
new	O
tricks	O
,	O
including	O
a	O
way	O
to	O
gin	O
user	O
ratings	O
.	O

DAN	O
GOODIN	O
-	O
1/23/2017	O
,	O
4:39	O
PM	O
A	O
virulent	B-Malware
family	O
of	O
malware	O
that	O
infected	O
more	O
than	O
10	O
million	O
Android	B-System
devices	O
last	O
year	O
has	O
made	O
a	O
comeback	O
,	O
this	O
time	O
hiding	O
inside	O
Google	B-System
Play	I-System
apps	O
that	O
have	O
been	O
downloaded	O
by	O
as	O
many	O
as	O
12	O
million	O
unsuspecting	O
users	O
.	O

HummingWhale	B-Malware
,	O
as	O
the	O
professionally	O
developed	O
malware	O
has	O
been	O
dubbed	O
,	O
is	O
a	O
variant	O
of	O
HummingBad	B-Malware
,	O
the	O
name	O
given	O
to	O
a	O
family	O
of	O
malicious	O
apps	O
researchers	O
documented	O
in	O
July	O
invading	O
non-Google	O
app	O
markets	O
.	O

HummingBad	B-Malware
attempted	O
to	O
override	O
security	O
protections	O
by	O
exploiting	O
unpatched	B-Vulnerability
vulnerabilities	I-Vulnerability
that	O
gave	O
the	O
malware	O
root	O
privileges	O
in	O
older	O
versions	O
of	O
Android	B-System
.	O

Before	O
Google	B-Organization
shut	O
it	O
down	O
,	O
it	O
installed	O
more	O
than	O
50,000	O
fraudulent	O
apps	O
each	O
day	O
,	O
displayed	O
20	O
million	O
malicious	O
advertisements	O
,	O
and	O
generated	O
more	O
than	O
$	O
300,000	O
per	O
month	O
in	O
revenue	O
.	O

Of	O
the	O
10	O
million	O
people	O
who	O
downloaded	O
HummingBad-contaminated	B-Malware
apps	O
,	O
an	O
estimated	O
286,000	O
of	O
them	O
were	O
located	O
in	O
the	O
US	O
.	O

HummingWhale	B-Malware
,	O
by	O
contrast	O
,	O
managed	O
to	O
sneak	O
its	O
way	O
into	O
about	O
20	O
Google	B-System
Play	I-System
apps	O
that	O
were	O
downloaded	O
from	O
2	O
million	O
to	O
12	O
million	O
times	O
,	O
according	O
to	O
researchers	O
from	O
Check	B-Organization
Point	I-Organization
,	O
the	O
security	O
company	O
that	O
has	O
been	O
closely	O
following	O
the	O
malware	O
family	O
for	O
almost	O
a	O
year	O
.	O

Rather	O
than	O
rooting	O
devices	O
,	O
the	O
latest	O
variant	O
includes	O
new	O
virtual	O
machine	O
techniques	O
that	O
allow	O
the	O
malware	O
to	O
perform	O
ad	O
fraud	O
better	O
than	O
ever	O
,	O
company	O
researchers	O
said	O
in	O
a	O
blog	O
post	O
published	O
Monday	O
.	O

"	O
Users	O
must	O
realize	O
that	O
they	O
can	O
no	O
longer	O
trust	O
in	O
installing	O
only	O
apps	O
with	O
a	O
high	O
reputation	O
from	O
official	O
app	O
stores	O
as	O
their	O
sole	O
defense	O
,	O
''	O
the	O
researchers	O
wrote	O
in	O
an	O
e-mail	O
to	O
Ars	B-Organization
.	O

"	O
This	O
malware	O
employs	O
several	O
tactics	O
to	O
keep	O
its	O
activity	O
hidden	O
,	O
meaning	O
users	O
might	O
be	O
unaware	O
of	O
its	O
existence	O
on	O
their	O
device	O
.	O

''	O
As	O
was	O
the	O
case	O
with	O
HummingBad	B-Malware
,	O
the	O
purpose	O
of	O
HummingWhale	B-Malware
is	O
to	O
generate	O
revenue	O
by	O
displaying	O
fraudulent	O
ads	O
and	O
automatically	O
installing	O
apps	O
.	O

When	O
users	O
try	O
to	O
close	O
the	O
ads	O
,	O
the	O
new	O
functionality	O
causes	O
already	O
downloaded	O
apps	O
to	O
run	O
in	O
a	O
virtual	O
machine	O
.	O

That	O
creates	O
a	O
fake	O
ID	O
that	O
allows	O
the	O
perpetrators	O
to	O
generate	O
referral	O
revenues	O
.	O

Use	O
of	O
the	O
virtual	O
machine	O
brings	O
many	O
technical	O
benefits	O
to	O
the	O
operators	O
,	O
chief	O
among	O
them	O
allowing	O
the	O
malware	O
to	O
install	O
apps	O
without	O
requiring	O
users	O
to	O
approve	O
a	O
list	O
of	O
elevated	O
permissions	O
.	O

Advertisement	O
The	O
VM	O
also	O
disguises	O
the	O
malicious	O
activity	O
,	O
making	O
it	O
easier	O
for	O
the	O
apps	O
to	O
infiltrate	O
Google	B-System
Play	I-System
.	O

It	O
has	O
the	O
added	O
benefit	O
of	O
installing	O
a	O
nearly	O
unlimited	O
number	O
of	O
fraudulent	O
apps	O
without	O
overloading	O
the	O
infected	O
device	O
.	O

Until	O
now	O
,	O
Android	B-System
malware	O
that	O
wanted	O
advanced	O
capabilities	O
typically	O
had	O
to	O
trick	O
users	O
into	O
approving	O
sometimes	O
scary-sounding	O
permissions	O
or	O
exploit	O
rooting	O
vulnerabilities	O
.	O

Ginning	O
the	O
ratings	O
FURTHER	O
READING	O
1	O
million	O
Google	B-Organization
accounts	O
compromised	O
by	O
Android	B-System
malware	O
called	O
Gooligan	B-Malware
To	O
implement	O
the	O
VM	O
feature	O
,	O
the	O
malicious	O
APK	O
installation	O
dropper	O
used	O
by	O
HummingWhale	B-Malware
uses	O
DroidPlugin	B-Malware
,	O
an	O
extension	O
originally	O
developed	O
by	O
developers	O
from	O
China-based	O
company	O
Qihoo	B-Organization
360	I-Organization
,	O
Check	B-Organization
Point	I-Organization
said	O
.	O

HummingWhale	B-Malware
has	O
also	O
been	O
observed	O
hiding	O
the	O
original	O
malicious	O
app	O
once	O
it	O
's	O
installed	O
and	O
trying	O
to	O
improve	O
its	O
Google	B-System
Play	I-System
reputation	O
by	O
automatically	O
generating	O
posts	O
disguised	O
as	O
positive	O
user	O
comments	O
and	O
ratings	O
.	O

Gooligan	B-Malware
,	O
a	O
family	O
of	O
Android	B-System
malware	O
that	O
came	O
to	O
light	O
in	O
November	O
after	O
it	O
compromised	O
more	O
than	O
1	O
million	O
Google	B-Organization
accounts	O
,	O
contained	O
similar	O
abilities	O
to	O
tamper	O
with	O
Google	B-System
Play	I-System
ratings	O
.	O

People	O
who	O
want	O
to	O
know	O
if	O
their	O
Android	B-System
devices	O
are	O
infected	O
can	O
download	O
the	O
Check	B-Organization
Point	I-Organization
app	O
here	O
.	O

A	O
separate	O
app	O
from	O
Check	B-Organization
Point	I-Organization
competitor	O
Lookout	B-Organization
also	O
detects	O
the	O
threat	O
as	O
a	O
variant	O
of	O
the	O
Shedun	B-Malware
malware	O
family	O
.	O

More	O
technically	O
inclined	O
people	O
can	O
detect	O
infections	O
by	O
seeing	O
if	O
a	O
device	O
connects	O
to	O
a	O
control	O
server	O
located	O
at	O
app.blinkingcamera.com	B-Indicator
.	O

Package	O
names	O
for	O
infected	O
apps	O
typically	O
contain	O
a	O
common	O
naming	O
structure	O
that	O
includes	O
com.XXXXXXXXX.camera	B-Indicator
,	O
for	O
example	O
com.bird.sky.whale.camera	B-Indicator
(	O
app	O
name	O
:	O
Whale	B-System
Camera	I-System
)	O
,	O
com.color.rainbow.camera	B-Indicator
(	O
Rainbow	B-System
Camera	I-System
)	O
,	O
and	O
com.fishing.when.orangecamera	B-Indicator
(	O
Orange	B-System
Camera	I-System
)	O
.	O

Google	B-Organization
officials	O
removed	O
the	O
malicious	O
apps	O
from	O
the	O
Play	B-System
market	I-System
after	O
receiving	O
a	O
private	O
report	O
of	O
their	O
existence	O
.	O

A	O
company	O
representative	O
declined	O
to	O
comment	O
for	O
this	O
post	O
.	O

BusyGasper	B-Malware
–	O
the	O
unfriendly	O
spy	O
29	O
AUG	O
2018	O
In	O
early	O
2018	O
our	O
mobile	O
intruder-detection	O
technology	O
was	O
triggered	O
by	O
a	O
suspicious	O
Android	B-System
sample	O
that	O
,	O
as	O
it	O
turned	O
out	O
,	O
belonged	O
to	O
an	O
unknown	O
spyware	O
family	O
.	O

Further	O
investigation	O
showed	O
that	O
the	O
malware	O
,	O
which	O
we	O
named	O
BusyGasper	B-Malware
,	O
is	O
not	O
all	O
that	O
sophisticated	O
,	O
but	O
demonstrates	O
some	O
unusual	O
features	O
for	O
this	O
type	O
of	O
threat	O
.	O

From	O
a	O
technical	O
point	O
of	O
view	O
,	O
the	O
sample	O
is	O
a	O
unique	O
spy	O
implant	O
with	O
stand-out	O
features	O
such	O
as	O
device	O
sensors	O
listeners	O
,	O
including	O
motion	O
detectors	O
that	O
have	O
been	O
implemented	O
with	O
a	O
degree	O
of	O
originality	O
.	O

It	O
has	O
an	O
incredibly	O
wide-ranging	O
protocol	O
–	O
about	O
100	O
commands	O
–	O
and	O
an	O
ability	O
to	O
bypass	O
the	O
Doze	O
battery	O
saver	O
.	O

As	O
a	O
modern	O
Android	O
spyware	O
it	O
is	O
also	O
capable	O
of	O
exfiltrating	O
data	O
from	O
messaging	O
applications	O
(	O
WhatsApp	B-System
,	O
Viber	B-System
,	O
Facebook	B-System
)	O
.	O

Moreover	O
,	O
BusyGasper	B-Malware
boasts	O
some	O
keylogging	O
tools	O
–	O
the	O
malware	O
processes	O
every	O
user	O
tap	O
,	O
gathering	O
its	O
coordinates	O
and	O
calculating	O
characters	O
by	O
matching	O
given	O
values	O
with	O
hardcoded	O
ones	O
.	O

The	O
sample	O
has	O
a	O
multicomponent	O
structure	O
and	O
can	O
download	O
a	O
payload	O
or	O
updates	O
from	O
its	O
C	O
&	O
C	O
server	O
,	O
which	O
happens	O
to	O
be	O
an	O
FTP	O
server	O
belonging	O
to	O
the	O
free	O
Russian	O
web	O
hosting	O
service	O
Ucoz	O
.	O

It	O
is	O
noteworthy	O
that	O
BusyGasper	B-Malware
supports	O
the	O
IRC	O
protocol	O
which	O
is	O
rarely	O
seen	O
among	O
Android	B-System
malware	O
.	O

In	O
addition	O
,	O
the	O
malware	O
can	O
log	O
in	O
to	O
the	O
attacker	O
’	O
s	O
email	O
inbox	O
,	O
parse	O
emails	O
in	O
a	O
special	O
folder	O
for	O
commands	O
and	O
save	O
any	O
payloads	O
to	O
a	O
device	O
from	O
email	O
attachments	O
.	O

This	O
particular	O
operation	O
has	O
been	O
active	O
since	O
approximately	O
May	O
2016	O
up	O
to	O
the	O
present	O
time	O
.	O

Infection	O
vector	O
and	O
victims	O
While	O
looking	O
for	O
the	O
infection	O
vector	O
,	O
we	O
found	O
no	O
evidence	O
of	O
spear	O
phishing	O
or	O
any	O
of	O
the	O
other	O
common	O
vectors	O
.	O

But	O
some	O
clues	O
,	O
such	O
as	O
the	O
existence	O
of	O
a	O
hidden	O
menu	O
for	O
operator	O
control	O
,	O
point	O
to	O
a	O
manual	O
installation	O
method	O
–	O
the	O
attackers	O
used	O
physical	O
access	O
to	O
a	O
victim	O
’	O
s	O
device	O
to	O
install	O
the	O
malware	O
.	O

This	O
would	O
explain	O
the	O
number	O
of	O
victims	O
–	O
there	O
are	O
less	O
than	O
10	O
of	O
them	O
and	O
according	O
to	O
our	O
detection	O
statistics	O
,	O
they	O
are	O
all	O
located	O
in	O
the	O
Russia	O
.	O

Intrigued	O
,	O
we	O
continued	O
our	O
search	O
and	O
found	O
more	O
interesting	O
clues	O
that	O
could	O
reveal	O
some	O
detailed	O
information	O
about	O
the	O
owners	O
of	O
the	O
infected	O
devices	O
.	O

Several	O
TXT	O
files	O
with	O
commands	O
on	O
the	O
attacker	O
’	O
s	O
FTP	O
server	O
contain	O
a	O
victim	O
identifier	O
in	O
the	O
names	O
that	O
was	O
probably	O
added	O
by	O
the	O
criminals	O
:	O
CMDS10114-Sun1.txt	B-Indicator
CMDS10134-Ju_ASUS.txt	B-Indicator
CMDS10134-Tad.txt	B-Indicator
CMDS10166-Jana.txt	B-Indicator
CMDS10187-Sun2.txt	B-Indicator
CMDS10194-SlavaAl.txt	B-Indicator
CMDS10209-Nikusha.txt	B-Indicator
Some	O
of	O
them	O
sound	O
like	O
Russian	O
names	O
:	O
Jana	O
,	O
SlavaAl	O
,	O
Nikusha	O
.	O

As	O
we	O
know	O
from	O
the	O
FTP	O
dump	O
analysis	O
,	O
there	O
was	O
a	O
firmware	O
component	O
from	O
ASUS	B-Organization
firmware	O
,	O
indicating	O
the	O
attacker	O
’	O
s	O
interest	O
in	O
ASUS	B-Organization
devices	O
,	O
which	O
explains	O
the	O
victim	O
file	O
name	O
that	O
mentions	O
“	O
ASUS	O
”	O
.	O

Information	O
gathered	O
from	O
the	O
email	O
account	O
provides	O
a	O
lot	O
of	O
the	O
victims	O
’	O
personal	O
data	O
,	O
including	O
messages	O
from	O
IM	O
applications	O
.	O

Gathered	O
file	O
Type	O
Description	O
lock	O
Text	O
Implant	O
log	O
ldata	O
sqlite3	O
Location	O
data	O
based	O
on	O
network	O
(	O
cell_id	O
)	O
gdata	O
sqlite3	O
Location	O
data	O
based	O
on	O
GPS	O
coordinates	O
sdata	B-Indicator
sqlite3	I-Indicator
SMS	O
messages	O
f.db	B-Indicator
sqlite3	I-Indicator
Facebook	B-System
messages	O
v.db	B-Indicator
sqlite3	I-Indicator
Viber	B-System
messages	O
w.db	B-Indicator
sqlite3	I-Indicator
WhatsApp	B-System
messages	O
Among	O
the	O
other	O
data	O
gathered	O
were	O
SMS	O
banking	O
messages	O
that	O
revealed	O
an	O
account	O
with	O
a	O
balance	O
of	O
more	O
than	O
US	O
$	O
10,000.But	O
as	O
far	O
as	O
we	O
know	O
,	O
the	O
attacker	O
behind	O
this	O
campaign	O
is	O
not	O
interested	O
in	O
stealing	O
the	O
victims	O
’	O
money	O

.	O

We	O
found	O
no	O
similarities	O
to	O
commercial	O
spyware	O
products	O
or	O
to	O
other	O
known	O
spyware	O
variants	O
,	O
which	O
suggests	O
BusyGasper	B-Malware
is	O
self-developed	O
and	O
used	O
by	O
a	O
single	O
threat	O
actor	O
.	O

At	O
the	O
same	O
time	O
,	O
the	O
lack	O
of	O
encryption	O
,	O
use	O
of	O
a	O
public	O
FTP	O
server	O
and	O
the	O
low	O
opsec	O
level	O
could	O
indicate	O
that	O
less	O
skilled	O
attackers	O
are	O
behind	O
the	O
malware	O
.	O

Technical	O
details	O
Here	O
is	O
the	O
meta	O
information	O
for	O
the	O
observed	O
samples	O
,	O
certificates	O
and	O
hardcoded	O
version	O
stamps	O
:	O
Certificate	O
MD5	O
Module	O
Version	O
Serial	O
Number	O
:	O
0x76607c02	B-Indicator
Issuer	O
:	O
CN=Ron	O
Validity	O
:	O
from	O
=	O
Tue	O
Aug	O
30	O
13:01:30	O
MSK	O
2016	O
to	O
=	O
Sat	O
Aug	O
24	O
13:01:30	O
MSK	O
2041	O
Subject	O
:	O
CN=Ron	O
9e005144ea1a583531f86663a5f14607	B-Indicator
1	O
–	O
18abe28730c53de6d9e4786c7765c3d8	B-Indicator
2	O
2.0	O

Serial	O
Number	O
:	O
0x6a0d1fec	B-Indicator
Issuer	O
:	O
CN=Sun	O
Validity	O
:	O
from	O
=	O
Mon	O
May	O
16	O
17:42:40	O
MSK	O
2016	O
to	O
=	O
Fri	O
May	O
10	O
17:42:40	O
MSK	O
2041	O
Subject	O
:	O
CN=Sun	O
9ffc350ef94ef840728564846f2802b0	B-Indicator
2	O
v2.51sun	O
6c246bbb40b7c6e75c60a55c0da9e2f2	B-Indicator
2	O
v2.96s	O
7c8a12e56e3e03938788b26b84b80bd6	B-Indicator
2	O
v3.09s	O

bde7847487125084f9e03f2b6b05adc3	B-Indicator
2	O
v3.12s	O
2560942bb50ee6e6f55afc495d238a12	B-Indicator
2	O
v3.18s	O
It	O
’	O
s	O
interesting	O
that	O
the	O
issuer	O
“	O
Sun	O
”	O
matches	O
the	O
“	O
Sun1	O
”	O
and	O
“	O
Sun2	O
”	O
identifiers	O
of	O
infected	O
devices	O
from	O
the	O
FTP	O
server	O
,	O
suggesting	O
they	O
may	O
be	O
test	O
devices	O
.	O

The	O
analyzed	O
implant	O
has	O
a	O
complex	O
structure	O
,	O
and	O
for	O
now	O
we	O
have	O
observed	O
two	O
modules	O
.	O

First	O
(	O
start	O
)	O
module	O
The	O
first	O
module	O
,	O
which	O
was	O
installed	O
on	O
the	O
targeted	O
device	O
,	O
could	O
be	O
controlled	O
over	O
the	O
IRC	O
protocol	O
and	O
enable	O
deployment	O
of	O
other	O
components	O
by	O
downloading	O
a	O
payload	O
from	O
the	O
FTP	O
server	O
:	O
@	O
install	O
command	O
As	O
can	O
be	O
seen	O
from	O
the	O
screenshot	O
above	O
,	O
a	O
new	O
component	O
was	O
copied	O
in	O
the	O
system	O
path	O
,	O
though	O
that	O
sort	O
of	O
operation	O
is	O
impossible	O
without	O
root	O
privileges	O
.	O

At	O
the	O
time	O
of	O
writing	O
we	O
had	O
no	O
evidence	O
of	O
an	O
exploit	O
being	O
used	O
to	O
obtain	O
root	O
privileges	O
,	O
though	O
it	O
is	O
possible	O
that	O
the	O
attackers	O
used	O
some	O
unseen	O
component	O
to	O
implement	O
this	O
feature	O
.	O

Here	O
is	O
a	O
full	O
list	O
of	O
possible	O
commands	O
that	O
can	O
be	O
executed	O
by	O
the	O
first	O
module	O
:	O
Command	O
name	O
Description	O
@	O
stop	O
Stop	O
IRC	O
@	O
quit	O
System.exit	B-Indicator
(	I-Indicator
0	I-Indicator
)	I-Indicator
@	O
start	O
Start	O
IRC	O
@	O
server	O
Set	O
IRC	O
server	O
(	O
default	O
value	O
is	O
“	O
irc.freenode.net	B-Indicator
”	O
)	O
,	O
port	O
is	O
always	O
6667	O
@	O
boss	O
Set	O
IRC	O
command	O
and	O
control	O
nickname	O
(	O
default	O
value	O
is	O
“	O
ISeency	O
”	O
)	O
@	O
nick	O
Set	O
IRC	O
client	O
nickname	O
@	O
screen	O
Report	O
every	O
time	O
when	O
screen	O
is	O
on	O
(	O
enable/disable	O
)	O
@	O
root	O
Use	O
root	O
features	O
(	O
enable/disable	O
)	O
@	O
timer	O
Set	O

period	O
of	O
IRCService	O
start	O
@	O
hide	O
Hide	O
implant	O
icon	O
@	O
unhide	O
Unhide	O
implant	O
icon	O
@	O
run	O
Execute	O
specified	O
shell	O
@	O
broadcast	O
Send	O
command	O
to	O
the	O
second	O
module	O
@	O
echo	O
Write	O
specified	O
message	O
to	O
log	O
@	O
install	O
Download	O
and	O
copy	O
specified	O
component	O
to	O
the	O
system	O
path	O
The	O
implant	O
uses	O
a	O
complex	O
intent-based	O
communication	O
mechanism	O
between	O
its	O
components	O
to	O
broadcast	O
commands	O
:	O
Approximate	O
graph	O
of	O
relationships	O
between	O
BusyGasper	O
components	O
Second	O
(	O
main	O
)	O
module	O
This	O
module	O
writes	O
a	O
log	O
of	O
the	O
command	O
execution	O
history	O
to	O
the	O
file	O
named	O
“	O
lock	O
”	O
,	O
which	O
is	O
later	O
exfiltrated	O

.	O

Below	O
is	O
a	O
fragment	O
of	O
such	O
a	O
log	O
:	O
Log	O
with	O
specified	O
command	O
Log	O
files	O
can	O
be	O
uploaded	O
to	O
the	O
FTP	O
server	O
and	O
sent	O
to	O
the	O
attacker	O
’	O
s	O
email	O
inbox	O
.	O

It	O
’	O
s	O
even	O
possible	O
to	O
send	O
log	O
messages	O
via	O
SMS	O
to	O
the	O
attacker	O
’	O
s	O
number	O
.	O

As	O
the	O
screenshot	O
above	O
shows	O
,	O
the	O
malware	O
has	O
its	O
own	O
command	O
syntax	O
that	O
represents	O
a	O
combination	O
of	O
characters	O
while	O
the	O
“	O
#	O
”	O
symbol	O
is	O
a	O
delimiter	O
.	O

A	O
full	O
list	O
of	O
all	O
possible	O
commands	O
with	O
descriptions	O
can	O
be	O
found	O
in	O
Appendix	O
II	O
below	O
.	O

The	O
malware	O
has	O
all	O
the	O
popular	O
capabilities	O
of	O
modern	O
spyware	O
.	O

Below	O
is	O
a	O
description	O
of	O
the	O
most	O
noteworthy	O
:	O
The	O
implant	O
is	O
able	O
to	O
spy	O
on	O
all	O
available	O
device	O
sensors	O
and	O
to	O
log	O
registered	O
events	O
.	O

Moreover	O
,	O
there	O
is	O
a	O
special	O
handler	O
for	O
the	O
accelerometer	O
that	O
is	O
able	O
to	O
calculate	O
and	O
log	O
the	O
device	O
’	O
s	O
speed	O
:	O
This	O
feature	O
is	O
used	O
in	O
particular	O
by	O
the	O
command	O
“	O
tk0	O
”	O
that	O
mutes	O
the	O
device	O
,	O
disables	O
keyguard	O
,	O
turns	O
off	O
the	O
brightness	O
,	O
uses	O
wakelock	O
and	O
listens	O
to	O
device	O
sensors	O
.	O

This	O
allows	O
it	O
to	O
silently	O
execute	O
any	O
backdoor	O
activity	O
without	O
the	O
user	O
knowing	O
that	O
the	O
device	O
is	O
in	O
an	O
active	O
state	O
.	O

As	O
soon	O
as	O
the	O
user	O
picks	O
up	O
the	O
device	O
,	O
the	O
implant	O
will	O
detect	O
a	O
motion	O
event	O
and	O
execute	O
the	O
“	O
tk1	O
”	O
and	O
“	O
input	O
keyevent	O
3	O
”	O
commands	O
.	O

“	O
tk1	O
”	O
will	O
disable	O
all	O
the	O
effects	O
of	O
the	O
“	O
tk0	O
”	O
command	O
,	O
while	O
“	O
input	O
keyevent	O
3	O
”	O
is	O
the	O
shell	O
command	O
that	O
simulates	O
the	O
pressing	O
of	O
the	O
‘	O
home	O
’	O
button	O
so	O
all	O
the	O
current	O
activities	O
will	O
be	O
minimized	O
and	O
the	O
user	O
won	O
’	O
t	O
suspect	O
anything	O
.	O

Location	O
services	O
to	O
enable	O
(	O
GPS/network	O
)	O
tracking	O
:	O
The	O
email	O
command	O
and	O
control	O
protocol	O
.	O

The	O
implant	O
can	O
log	O
in	O
to	O
the	O
attackers	O
email	O
inbox	O
,	O
parse	O
emails	O
for	O
commands	O
in	O
a	O
special	O
“	O
Cmd	O
”	O
folder	O
and	O
save	O
any	O
payloads	O
to	O
a	O
device	O
from	O
email	O
attachments	O
.	O

Accessing	O
the	O
“	O
Cmd	O
”	O
folder	O
in	O
the	O
attacker	O
’	O
s	O
email	O
box	O
Moreover	O
,	O
it	O
can	O
send	O
a	O
specified	O
file	O
or	O
all	O
the	O
gathered	O
data	O
from	O
the	O
victim	O
device	O
via	O
email	O
.	O

Emergency	O
SMS	O
commands	O
.	O

If	O
an	O
incoming	O
SMS	O
contains	O
one	O
of	O
the	O
following	O
magic	O
strings	O
:	O
”	O
2736428734″	B-Indicator
or	O
”	O
7238742800″	B-Indicator
the	O
malware	O
will	O
execute	O
multiple	O
initial	O
commands	O
:	O
Keylogger	O
implementation	O
Keylogging	O
is	O
implemented	O
in	O
an	O
original	O
manner	O
.	O

Immediately	O
after	O
activation	O
,	O
the	O
malware	O
creates	O
a	O
textView	O
element	O
in	O
a	O
new	O
window	O
with	O
the	O
following	O
layout	O
parameters	O
:	O
All	O
these	O
parameters	O
ensure	O
the	O
element	O
is	O
hidden	O
from	O
the	O
user	O
.	O

Then	O
it	O
adds	O
onTouchListener	O
to	O
this	O
textView	O
and	O
is	O
able	O
to	O
process	O
every	O
user	O
tap	O
.	O

Interestingly	O
,	O
there	O
is	O
an	O
allowlist	O
of	O
tapped	O
activities	O
:	O
ui.ConversationActivity	O
ui.ConversationListActivity	O
SemcInCallScreen	O
Quadrapop	O
SocialPhonebookActivity	O
The	O
listener	O
can	O
operate	O
with	O
only	O
coordinates	O
,	O
so	O
it	O
calculates	O
pressed	O
characters	O
by	O
matching	O
given	O
values	O
with	O
hardcoded	O
ones	O
:	O
Additionally	O
,	O
if	O
there	O
is	O
a	O
predefined	O
command	O
,	O
the	O
keylogger	O
can	O
make	O
a	O
screenshot	O
of	O
the	O
tapped	O
display	O
area	O
:	O
Manual	O
access	O
and	O
operator	O
menu	O
There	O
is	O
a	O
hidden	O
menu	O
(	O
Activity	O
)	O
for	O
controlling	O
implant	O
features	O
that	O

looks	O
like	O
it	O
was	O
created	O
for	O
manual	O
operator	O
control	O
.	O

To	O
activate	O
this	O
menu	O
the	O
operator	O
needs	O
to	O
call	O
the	O
hardcoded	O
number	O
“	O
9909	O
”	O
from	O
the	O
infected	O
device	O
:	O
A	O
hidden	O
menu	O
then	O
instantly	O
appears	O
on	O
the	O
device	O
display	O
:	O
The	O
operator	O
can	O
use	O
this	O
interface	O
to	O
type	O
any	O
command	O
for	O
execution	O
.	O

It	O
also	O
shows	O
a	O
current	O
malware	O
log	O
.	O

Infrastructure	O
FTP	O
server	O
The	O
attackers	O
used	O
ftp	B-Indicator
:	I-Indicator
//213.174.157	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
151/	I-Indicator
as	O
a	O
command	O
and	O
control	O
server	O
.	O

The	O
IP	O
belongs	O
to	O
the	O
free	O
Russian	O
web	O
hosting	O
service	O
Ucoz	O
.	O

Files	O
Description	O
CMDS	O
*	O
.txt	O
Text	O
files	O
with	O
commands	O
to	O
execute	O
supersu.apk	B-Indicator
SuperSU	O
(	O
eu.chainfire.supersu	B-Indicator
,	O
https	B-Indicator
:	I-Indicator
//play.google.com/store/apps/details	I-Indicator
?	I-Indicator

id=eu.chainfire.supersu	I-Indicator
)	O
tool	O
246.us	B-Indicator
us.x	B-Indicator
SuperSU	O
ELF	O
binaries	O
supersu.cfg	B-Indicator
supersu.cfg.ju	B-Indicator
supersu.cfg.old	B-Indicator
SuperSU	O
configs	O
with	O
spyware	O
implant	O
mention	O
bb.txt	B-Indicator
BusyBox	O
v1.26.2	O
ELF	O
file	O
bdata.xml	B-Indicator
Config	O
file	O
for	O
excluding	O
malware	O
components	O
from	O
Android	B-System
battery	O
saver	O
feature	O
Doze	O
bdatas.apk	B-Indicator
Main	O
implant	O
module	O
com.android.network.irc.apk	B-Indicator
Start	O
implant	O
module	O
MobileManagerService.apk	B-Indicator
ASUS	B-Organization
firmware	O
system	O
component	O
(	O
clean	O
)	O
mobilemanager.apk	B-Indicator

Corrupted	O
archive	O
privapp.txt	B-Indicator
Looks	O
like	O
a	O
list	O
of	O
system	O
applications	O
(	O
including	O
spyware	O
components	O
)	O
from	O
the	O
infected	O
device	O
run-as.x	B-Indicator
run-as.y	B-Indicator
Run-as	O
tool	O
ELF	O
file	O
SuperSU	O
config	O
fragment	O
for	O
implant	O
components	O
and	O
the	O
busybox	O
tool	O
supersu.cfg	B-Indicator
:	O
This	O
config	O
allows	O
the	O
implant	O
to	O
use	O
all	O
root	O
features	O
silently	O
.	O

Content	O
of	O
bdata.xml	O
file	O
:	O
It	O
can	O
be	O
added	O
to	O
the	O
/system/etc/sysconfig/	B-Indicator
path	O
to	O
allowlist	O
specified	O
implant	O
components	O
from	O
the	O
battery	O
saving	O
system	O
.	O

Email	O
account	O
A	O
Gmail	B-System
account	O
with	O
password	O
is	O
mentioned	O
in	O
the	O
sample	O
’	O
s	O
code	O
:	O
It	O
contains	O
the	O
victim	O
’	O
s	O
exfiltrated	O
data	O
and	O
“	O
cmd	O
”	O
directory	O
with	O
commands	O
for	O
victim	O
devices	O
.	O

10	O
million	O
Android	B-System
phones	O
infected	O
by	O
all-powerful	O
auto-rooting	O
apps	O
First	O
detected	O
in	O
November	O
,	O
Shedun/HummingBad	B-Malware
infections	O
are	O
surging	O
.	O

7/7/2016	O
,	O
1:50	O
PM	O
Security	O
experts	O
have	O
documented	O
a	O
disturbing	O
spike	O
in	O
a	O
particularly	O
virulent	O
family	O
of	O
Android	B-System
malware	O
,	O
with	O
more	O
than	O
10	O
million	O
handsets	O
infected	O
and	O
more	O
than	O
286,000	O
of	O
them	O
in	O
the	O
US	O
.	O

FURTHER	O
READING	O
New	O
type	O
of	O
auto-rooting	O
Android	B-System
adware	O
is	O
nearly	O
impossible	O
to	O
remove	O
Researchers	O
from	O
security	O
firm	O
Check	B-Organization
Point	I-Organization
Software	I-Organization
said	O
the	O
malware	O
installs	O
more	O
than	O
50,000	O
fraudulent	O
apps	O
each	O
day	O
,	O
displays	O
20	O
million	O
malicious	O
advertisements	O
,	O
and	O
generates	O
more	O
than	O
$	O
300,000	O
per	O
month	O
in	O
revenue	O
.	O

The	O
success	O
is	O
largely	O
the	O
result	O
of	O
the	O
malware	O
's	O
ability	O
to	O
silently	O
root	O
a	O
large	O
percentage	O
of	O
the	O
phones	O
it	O
infects	O
by	O
exploiting	O
vulnerabilities	B-Vulnerability
that	I-Vulnerability
remain	I-Vulnerability
unfixed	I-Vulnerability
in	I-Vulnerability
older	I-Vulnerability
versions	I-Vulnerability
of	I-Vulnerability
Android	I-Vulnerability
.	O

The	O
Check	B-Organization
Point	I-Organization
researchers	O
have	O
dubbed	O
the	O
malware	O
family	O
"	O
HummingBad	B-Malware
,	O
''	O
but	O
researchers	O
from	O
mobile	O
security	O
company	O
Lookout	B-Organization
say	O
HummingBad	B-Malware
is	O
in	O
fact	O
Shedun	B-Malware
,	O
a	O
family	O
of	O
auto-rooting	O
malware	O
that	O
came	O
to	O
light	O
last	O
November	O
and	O
had	O
already	O
infected	O
a	O
large	O
number	O
of	O
devices	O
.	O

Update	O
Jul	O
11	O
2016	O
8:32	O
:	O
On	O
Monday	O
,	O
a	O
Checkpoint	B-Organization
representative	O
disputed	O
Lookout	B-Organization
's	O
contention	O
and	O
pointed	O
to	O
this	O
blog	O
post	O
from	O
security	O
firm	O
Eleven	B-Organization
Paths	I-Organization
as	O
support	O
.	O

The	O
blog	O
post	O
said	O
HummingBad	B-Malware
"	O
uses	O
a	O
completely	O
different	O
infrastructure	O
with	O
little	O
in	O
common	O
''	O
with	O
Shedun	I-Malware
.	O

In	O
an	O
e-mail	O
,	O
a	O
Lookout	B-Organization
representative	O
stood	O
by	O
its	O
analysis	O
and	O
said	O
company	O
researchers	O
planned	O
to	O
publish	O
an	O
in-depth	O
response	O
in	O
the	O
coming	O
days	O
.	O

For	O
the	O
past	O
five	O
months	O
,	O
Check	B-Organization
Point	I-Organization
researchers	O
have	O
quietly	O
observed	O
the	O
China-based	O
advertising	O
company	O
behind	O
HummingBad	B-Malware
in	O
several	O
ways	O
,	O
including	O
by	O
infiltrating	O
the	O
command	O
and	O
control	O
servers	O
it	O
uses	O
.	O

The	O
researchers	O
say	O
the	O
malware	O
uses	O
the	O
unusually	O
tight	O
control	O
it	O
gains	O
over	O
infected	O
devices	O
to	O
create	O
windfall	O
profits	O
and	O
steadily	O
increase	O
its	O
numbers	O
.	O

HummingBad	B-Malware
does	O
this	O
by	O
silently	O
installing	O
promoted	O
apps	O
on	O
infected	O
phones	O
,	O
defrauding	O
legitimate	O
mobile	O
advertisers	O
,	O
and	O
creating	O
fraudulent	O
statistics	O
inside	O
the	O
official	O
Google	B-System
Play	I-System
Store	I-System
.	O

"	O
Accessing	O
these	O
devices	O
and	O
their	O
sensitive	O
data	O
creates	O
a	O
new	O
and	O
steady	O
stream	O
of	O
revenue	O
for	O
cybercriminals	O
,	O
''	O
Check	I-Organization
Point	I-Organization
researchers	O
wrote	O
in	O
a	O
recently	O
published	O
report	O
.	O

"	O
Emboldened	O
by	O
financial	O
and	O
technological	O
independence	O
,	O
their	O
skillsets	O
will	O
advance–putting	O
end	O
users	O
,	O
enterprises	O
,	O
and	O
government	O
agencies	O
at	O
risk	O
.	O

''	O
The	O
report	O
said	O
HummingBad	B-Malware
apps	O
are	O
developed	O
by	O
Yingmob	B-Organization
,	O
a	O
Chinese	O
mobile	O
ad	O
server	O
company	O
that	O
other	O
researchers	O
claim	O
is	O
behind	O
the	O
Yinspector	B-Malware
iOS	B-System
malware	O
.	O

HummingBad	B-Malware
sends	O
notifications	O
to	O
Umeng	O
,	O
a	O
tracking	O
and	O
analytics	O
service	O
attackers	O
use	O
to	O
manage	O
their	O
campaign	O
.	O

Check	B-Organization
Point	I-Organization
analyzed	O
Yingmob	B-Organization
’	O
s	O
Umeng	O
account	O
to	O
gain	O
further	O
insights	O
into	O
the	O
HummingBad	B-Malware
campaign	O
and	O
found	O
that	O
beyond	O
the	O
10	O
million	O
devices	O
under	O
the	O
control	O
of	O
malicious	O
apps	O
,	O
Yingmob	B-Organization
has	O
non-malicious	O
apps	O
installed	O
on	O
another	O
75	O
million	O
or	O
so	O
devices	O
.	O

The	O
researchers	O
wrote	O
:	O
While	O
profit	O
is	O
powerful	O
motivation	O
for	O
any	O
attacker	O
,	O
Yingmob	B-Organization
’	O
s	O
apparent	O
self-sufficiency	O
and	O
organizational	O
structure	O
make	O
it	O
well-positioned	O
to	O
expand	O
into	O
new	O
business	O
ventures	O
,	O
including	O
productizing	O
the	O
access	O
to	O
the	O
85	O
million	O
Android	B-System
devices	O
it	O
controls	O
.	O

This	O
alone	O
would	O
attract	O
a	O
whole	O
new	O
audience–and	O
a	O
new	O
stream	O
of	O
revenue–for	O
Yingmob	B-Organization
.	O

Quick	O
,	O
easy	O
access	O
to	O
sensitive	O
data	O
on	O
mobile	O
devices	O
connected	O
to	O
enterprises	O
and	O
government	O
agencies	O
around	O
the	O
globe	O
is	O
extremely	O
attractive	O
to	O
cybercriminals	O
and	O
hacktivists	O
.	O

Drive-by	O
downloads	O
and	O
multiple	O
rooting	O
exploits	O
The	O
malware	O
uses	O
a	O
variety	O
of	O
methods	O
to	O
infect	O
devices	O
.	O

One	O
involves	O
drive-by	O
downloads	O
,	O
possibly	O
on	O
booby-trapped	O
porn	O
sites	O
.	O

The	O
attacks	O
use	O
multiple	O
exploits	O
in	O
an	O
attempt	O
to	O
gain	O
root	O
access	O
on	O
a	O
device	O
.	O

When	O
rooting	O
fails	O
,	O
a	O
second	O
component	O
delivers	O
a	O
fake	O
system	O
update	O
notification	O
in	O
hopes	O
of	O
tricking	O
users	O
into	O
granting	O
HummingBad	B-Malware
system-level	O
permissions	O
.	O

Whether	O
or	O
not	O
rooting	O
succeeds	O
,	O
HummingBad	B-Malware
downloads	O
a	O
large	O
number	O
of	O
apps	O
.	O

In	O
some	O
cases	O
,	O
malicious	O
components	O
are	O
dynamically	O
downloaded	O
onto	O
a	O
device	O
after	O
an	O
infected	O
app	O
is	O
installed	O
.	O

From	O
there	O
,	O
infected	O
phones	O
display	O
illegitimate	O
ads	O
and	O
install	O
fraudulent	O
apps	O
after	O
certain	O
events	O
,	O
such	O
as	O
rebooting	O
,	O
the	O
screen	O
turning	O
on	O
or	O
off	O
,	O
a	O
detection	O
that	O
the	O
user	O
is	O
present	O
,	O
or	O
a	O
change	O
in	O
Internet	O
connectivity	O
.	O

HummingBad	B-Malware
also	O
has	O
the	O
ability	O
to	O
inject	O
code	O
into	O
Google	B-System
Play	I-System
to	O
tamper	O
with	O
its	O
ratings	O
and	O
statistics	O
.	O

It	O
does	O
this	O
by	O
using	O
infected	O
devices	O
to	O
imitate	O
clicks	O
on	O
the	O
install	O
,	O
buy	O
,	O
and	O
accept	O
buttons	O
.	O

Many	O
of	O
the	O
10	O
million	O
infected	O
phones	O
are	O
running	O
old	O
versions	O
of	O
Android	B-System
and	O
reside	O
in	O
China	O
(	O
1.6	O
million	O
)	O
and	O
India	O
(	O
1.35	O
million	O
)	O
.	O

Still	O
,	O
US-based	O
infected	O
phones	O
total	O
almost	O
287,000	O
.	O

The	O
most	O
widely	O
infected	O
major	O
Android	B-System
versions	O
are	O
KitKat	B-System
with	O
50	O
percent	O
,	O
followed	O
by	O
Jelly	B-System
Bean	I-System
with	O
40	O
percent	O
.	O

Lollipop	B-System
has	O
7	O
percent	O
,	O
Ice	B-System
Cream	I-System
Sandwich	I-System
has	O
2	O
percent	O
,	O
and	O
Marshmallow	B-System
has	O
1	O
percent	O
.	O

It	O
's	O
often	O
hard	O
for	O
average	O
users	O
to	O
know	O
if	O
their	O
phones	O
have	O
been	O
rooted	O
,	O
and	O
Shedun	B-Malware
apps	O
often	O
wait	O
some	O
period	O
of	O
time	O
before	O
displaying	O
obtrusive	O
ads	O
or	O
installing	O
apps	O
.	O

The	O
best	O
bet	O
for	O
Readers	O
who	O
want	O
to	O
make	O
sure	O
their	O
phone	O
is	O
n't	O
infected	O
is	O
to	O
scan	O
their	O
phones	O
using	O
the	O
free	O
version	O
of	O
the	O
Lookout	B-Organization
Security	O
and	O
Antivirus	O
app	O
.	O

Android	B-System
malware	O
has	O
drastically	O
lower	O
rates	O
of	O
success	O
when	O
app	O
installations	O
outside	O
of	O
Google	B-System
Play	I-System
are	O
barred	O
.	O

Readers	O
should	O
carefully	O
think	O
through	O
the	O
risks	O
before	O
changing	O
this	O
default	O
setting	O
.	O

Top	O
20	O
countries	O
targeted	O
by	O
Hummingbad/Shedun	B-Malware
.	O

Enlarge	O
/	O
Top	O
20	O
countries	O
targeted	O
by	O
Hummingbad/Shedun	B-Malware
.	O

Check	B-Organization
Point	I-Organization
Software	I-Organization
Hummingbad/Shedun	B-Malware
infections	O
by	O
Android	B-System
version	O
.	O

Enlarge	O
/	O
Hummingbad/Shedun	B-Malware
infections	O
by	O
Android	B-System
version	O
.	O

Check	B-Organization
Point	I-Organization
Software	I-Organization
So	O
far	O
,	O
HummingBad	B-Malware
has	O
been	O
observed	O
using	O
its	O
highly	O
privileged	O
status	O
only	O
to	O
engage	O
in	O
click	O
fraud	O
,	O
display	O
pop-up	O
ads	O
,	O
tamper	O
with	O
Google	B-System
Play	I-System
,	O
and	O
install	O
additional	O
apps	O
that	O
do	O
more	O
of	O
the	O
same	O
.	O

But	O
there	O
's	O
little	O
stopping	O
it	O
from	O
doing	O
much	O
worse	O
.	O

That	O
's	O
because	O
the	O
malware	O
roots	O
most	O
of	O
the	O
phones	O
it	O
infects	O
,	O
a	O
process	O
that	O
subverts	O
key	O
security	O
mechanisms	O
built	O
into	O
Android	B-System
.	O

Under	O
a	O
model	O
known	O
as	O
sandboxing	O
,	O
most	O
Android	B-System
apps	O
are	O
n't	O
permitted	O
to	O
access	O
passwords	O
or	O
other	O
data	O
available	O
to	O
most	O
other	O
apps	O
.	O

System	O
applications	O
with	O
root	O
,	O
by	O
contrast	O
,	O
have	O
super-user	O
permissions	O
that	O
allow	O
them	O
to	O
break	O
out	O
of	O
such	O
sandboxes	O
.	O

From	O
there	O
,	O
root-level	O
apps	O
can	O
read	O
or	O
modify	O
data	O
and	O
resources	O
that	O
would	O
be	O
off-limits	O
to	O
normal	O
apps	O
.	O

As	O
Lookout	B-Organization
first	O
reported	O
more	O
than	O
eight	O
months	O
ago	O
,	O
the	O
problem	O
with	O
Shedun/HummingBad	B-Malware
and	O
similar	O
malicious	O
app	O
families	O
that	O
silently	O
exploit	O
Android	B-Vulnerability
rooting	I-Vulnerability
vulnerabilities	I-Vulnerability
is	O
that	O
the	O
infections	O
can	O
survive	O
normal	O
factory	O
resets	O
.	O

Lookout	B-Organization
said	O
in	O
its	O
own	O
blog	O
post	O
published	O
Wednesday	O
that	O
its	O
threat	O
detection	O
network	O
has	O
recently	O
observed	O
a	O
surge	O
of	O
Shedun	B-Malware
attacks	O
,	O
indicating	O
the	O
scourge	O
wo	O
n't	O
be	O
going	O
away	O
any	O
time	O
soon	O
.	O

An	O
investigation	O
of	O
Chrysaor	B-Malware
Malware	O
on	O
Android	B-System
03	O
April	O
2017	O
Google	B-Organization
is	O
constantly	O
working	O
to	O
improve	O
our	O
systems	O
that	O
protect	O
users	O
from	O
Potentially	O
Harmful	O
Applications	O
(	O
PHAs	O
)	O
.	O

Usually	O
,	O
PHA	O
authors	O
attempt	O
to	O
install	O
their	O
harmful	O
apps	O
on	O
as	O
many	O
devices	O
as	O
possible	O
.	O

However	O
,	O
a	O
few	O
PHA	O
authors	O
spend	O
substantial	O
effort	O
,	O
time	O
,	O
and	O
money	O
to	O
create	O
and	O
install	O
their	O
harmful	O
app	O
on	O
one	O
or	O
a	O
very	O
small	O
number	O
of	O
devices	O
.	O

This	O
is	O
known	O
as	O
a	O
targeted	O
attack	O
.	O

In	O
this	O
blog	O
post	O
,	O
we	O
describe	O
Chrysaor	B-Malware
,	O
a	O
newly	O
discovered	O
family	O
of	O
spyware	O
that	O
was	O
used	O
in	O
a	O
targeted	O
attack	O
on	O
a	O
small	O
number	O
of	O
Android	B-System
devices	O
,	O
and	O
how	O
investigations	O
like	O
this	O
help	O
Google	B-Organization
protect	O
Android	B-System
users	O
from	O
a	O
variety	O
of	O
threats	O
.	O

What	O
is	O
Chrysaor	B-Malware
?	O

Chrysaor	B-Malware
is	O
spyware	O
believed	O
to	O
be	O
created	O
by	O
NSO	B-Organization
Group	I-Organization
Technologies	I-Organization
,	O
specializing	O
in	O
the	O
creation	O
and	O
sale	O
of	O
software	O
and	O
infrastructure	O
for	O
targeted	O
attacks	O
.	O

Chrysaor	B-Malware
is	O
believed	O
to	O
be	O
related	O
to	O
the	O
Pegasus	B-Malware
spyware	O
that	O
was	O
first	O
identified	O
on	O
iOS	B-System
and	O
analyzed	O
by	O
Citizen	B-Organization
Lab	I-Organization
and	O
Lookout	B-Organization
.	O

Late	O
last	O
year	O
,	O
after	O
receiving	O
a	O
list	O
of	O
suspicious	O
package	O
names	O
from	O
Lookout	B-Organization
,	O
we	O
discovered	O
that	O
a	O
few	O
dozen	O
Android	B-System
devices	O
may	O
have	O
installed	O
an	O
application	O
related	O
to	O
Pegasus	B-Malware
,	O
which	O
we	O
named	O
Chrysaor	B-Malware
.	O

Although	O
the	O
applications	O
were	O
never	O
available	O
in	O
Google	B-System
Play	I-System
,	O
we	O
immediately	O
identified	O
the	O
scope	O
of	O
the	O
problem	O
by	O
using	O
Verify	B-System
Apps	I-System
.	O

We	O
gathered	O
information	O
from	O
affected	O
devices	O
,	O
and	O
concurrently	O
,	O
attempted	O
to	O
acquire	O
Chrysaor	B-Malware
apps	O
to	O
better	O
understand	O
its	O
impact	O
on	O
users	O
.	O

We	O
've	O
contacted	O
the	O
potentially	O
affected	O
users	O
,	O
disabled	O
the	O
applications	O
on	O
affected	O
devices	O
,	O
and	O
implemented	O
changes	O
in	O
Verify	B-System
Apps	I-System
to	O
protect	O
all	O
users	O
.	O

What	O
is	O
the	O
scope	O
of	O
Chrysaor	B-Malware
?	O

Chrysaor	B-Malware
was	O
never	O
available	O
in	O
Google	B-System
Play	I-System
and	O
had	O
a	O
very	O
low	O
volume	O
of	O
installs	O
outside	O
of	O
Google	B-System
Play	I-System
.	O

Among	O
the	O
over	O
1.4	O
billion	O
devices	O
protected	O
by	O
Verify	B-System
Apps	I-System
,	O
we	O
observed	O
fewer	O
than	O
3	O
dozen	O
installs	O
of	O
Chrysaor	B-Malware
on	O
victim	O
devices	O
.	O

These	O
devices	O
were	O
located	O
in	O
the	O
following	O
countries	O
:	O
How	O
we	O
protect	O
you	O
To	O
protect	O
Android	B-System
devices	O
and	O
users	O
,	O
Google	B-System
Play	I-System
provides	O
a	O
complete	O
set	O
of	O
security	O
services	O
that	O
update	O
outside	O
of	O
platform	O
releases	O
.	O

Users	O
do	O
n't	O
have	O
to	O
install	O
any	O
additional	O
security	O
services	O
to	O
keep	O
their	O
devices	O
safe	O
.	O

In	O
2016	O
,	O
these	O
services	O
protected	O
over	O
1.4	O
billion	O
devices	O
,	O
making	O
Google	B-Organization
one	O
of	O
the	O
largest	O
providers	O
of	O
on-device	O
security	O
services	O
in	O
the	O
world	O
:	O
Identify	O
PHAs	O
using	O
people	O
,	O
systems	O
in	O
the	O
cloud	O
,	O
and	O
data	O
sent	O
to	O
us	O
from	O
devices	O
Warn	O
users	O
about	O
or	O
blocking	O
users	O
from	O
installing	O
PHAs	O
Continually	O
scan	O
devices	O
for	O
PHAs	O
and	O
other	O
harmful	O
threats	O
Additionally	O
,	O
we	O
are	O
providing	O
detailed	O
technical	O
information	O
to	O
help	O
the	O
security	O
industry	O
in	O
our	O
collective	O
work	O
against	O
PHAs	O
.	O

What	O
do	O
I	O
need	O
to	O
do	O
?	O

It	O
is	O
extremely	O
unlikely	O
you	O
or	O
someone	O
you	O
know	O
was	O
affected	O
by	O
Chrysaor	B-Malware
malware	O
.	O

Through	O
our	O
investigation	O
,	O
we	O
identified	O
less	O
than	O
3	O
dozen	O
devices	O
affected	O
by	O
Chrysaor	B-Malware
,	O
we	O
have	O
disabled	O
Chrysaor	B-Malware
on	O
those	O
devices	O
,	O
and	O
we	O
have	O
notified	O
users	O
of	O
all	O
known	O
affected	O
devices	O
.	O

Additionally	O
,	O
the	O
improvements	O
we	O
made	O
to	O
our	O
protections	O
have	O
been	O
enabled	O
for	O
all	O
users	O
of	O
our	O
security	O
services	O
.	O

To	O
ensure	O
you	O
are	O
fully	O
protected	O
against	O
PHAs	O
and	O
other	O
threats	O
,	O
we	O
recommend	O
these	O
5	O
basic	O
steps	O
:	O
Install	O
apps	O
only	O
from	O
reputable	O
sources	O
:	O
Install	O
apps	O
from	O
a	O
reputable	O
source	O
,	O
such	O
as	O
Google	B-System
Play	I-System
.	O

No	O
Chrysaor	B-Malware
apps	O
were	O
on	O
Google	B-System
Play	I-System
.	O

Enable	O
a	O
secure	O
lock	O
screen	O
:	O
Pick	O
a	O
PIN	O
,	O
pattern	O
,	O
or	O
password	O
that	O
is	O
easy	O
for	O
you	O
to	O
remember	O
and	O
hard	O
for	O
others	O
to	O
guess	O
.	O

Update	O
your	O
device	O
:	O
Keep	O
your	O
device	O
up-to-date	O
with	O
the	O
latest	O
security	O
patches	O
.	O

Verify	O
Apps	O
:	O
Ensure	O
Verify	O
Apps	O
is	O
enabled	O
.	O

Locate	O
your	O
device	O
:	O
Practice	O
finding	O
your	O
device	O
with	O
Android	B-System
Device	I-System
Manager	I-System
because	O
you	O
are	O
far	O
more	O
likely	O
to	O
lose	O
your	O
device	O
than	O
install	O
a	O
PHA	O
.	O

How	O
does	O
Chrysaor	B-Malware
work	O
?	O

To	O
install	O
Chrysaor	B-Malware
,	O
we	O
believe	O
an	O
attacker	O
coaxed	O
specifically	O
targeted	O
individuals	O
to	O
download	O
the	O
malicious	O
software	O
onto	O
their	O
device	O
.	O

Once	O
Chrysaor	B-Malware
is	O
installed	O
,	O
a	O
remote	O
operator	O
is	O
able	O
to	O
surveil	O
the	O
victim	O
's	O
activities	O
on	O
the	O
device	O
and	O
within	O
the	O
vicinity	O
,	O
leveraging	O
microphone	O
,	O
camera	O
,	O
data	O
collection	O
,	O
and	O
logging	O
and	O
tracking	O
application	O
activities	O
on	O
communication	O
apps	O
such	O
as	O
phone	O
and	O
SMS	O
.	O

One	O
representative	O
sample	O
Chrysaor	B-Malware
app	O
that	O
we	O
analyzed	O
was	O
tailored	O
to	O
devices	O
running	O
Jellybean	B-System
(	I-System
4.3	I-System
)	I-System
or	O
earlier	O
.	O

The	O
following	O
is	O
a	O
review	O
of	O
scope	O
and	O
impact	O
of	O
the	O
Chrysaor	B-Malware
app	O
named	O
com.network.android	B-Indicator
tailored	O
for	O
a	O
Samsung	B-Organization
device	O
target	O
,	O
with	O
SHA256	O
digest	O
:	O
ade8bef0ac29fa363fc9afd958af0074478aef650adeb0318517b48bd996d5d5Upon	B-Indicator
installation	O
,	O
the	O
app	O
uses	O
known	O
framaroot	O
exploits	O
to	O
escalate	O
privileges	O
and	O
break	O
Android	B-System
's	O
application	O
sandbox	O
.	O

If	O
the	O
targeted	O
device	O
is	O
not	O
vulnerable	O
to	O
these	O
exploits	O
,	O
then	O
the	O
app	O
attempts	O
to	O
use	O
a	O
superuser	O
binary	O
pre-positioned	O
at	O
/system/csk	B-Indicator
to	O
elevate	O
privileges	O
.	O

After	O
escalating	O
privileges	O
,	O
the	O
app	O
immediately	O
protects	O
itself	O
and	O
starts	O
to	O
collect	O
data	O
,	O
by	O
:	O
Installing	O
itself	O
on	O
the	O
/system	O
partition	O
to	O
persist	O
across	O
factory	O
resets	O
Removing	O
Samsung	B-Organization
's	O
system	O
update	O
app	O
(	O
com.sec.android.fotaclient	B-Indicator
)	O
and	O
disabling	O
auto-updates	O
to	O
maintain	O
persistence	O
(	O
sets	O
Settings.System.SOFTWARE_UPDATE_AUTO_UPDATE	B-Indicator
to	I-Indicator
0	I-Indicator
)	O
Deleting	O
WAP	O
push	O
messages	O
and	O
changing	O
WAP	O
message	O
settings	O
,	O
possibly	O
for	O
anti-forensic	O
purpose	O
.	O

Starting	O
content	O
observers	O
and	O
the	O
main	O
task	O
loop	O
to	O
receive	O
remote	O
commands	O
and	O
exfiltrate	O
data	O
The	O
app	O
uses	O
six	O
techniques	O
to	O
collect	O
user	O
data	O
:	O
Repeated	O
commands	O
:	O
use	O
alarms	O
to	O
periodically	O
repeat	O
actions	O
on	O
the	O
device	O
to	O
expose	O
data	O
,	O
including	O
gathering	O
location	O
data	O
.	O

Data	O
collectors	O
:	O
dump	O
all	O
existing	O
content	O
on	O
the	O
device	O
into	O
a	O
queue	O
.	O

Data	O
collectors	O
are	O
used	O
in	O
conjunction	O
with	O
repeated	O
commands	O
to	O
collect	O
user	O
data	O
including	O
,	O
SMS	O
settings	O
,	O
SMS	O
messages	O
,	O
Call	O
logs	O
,	O
Browser	O
History	O
,	O
Calendar	O
,	O
Contacts	O
,	O
Emails	O
,	O
and	O
messages	O
from	O
selected	O
messaging	O
apps	O
,	O
including	O
WhatsApp	B-System
,	O
Twitter	B-System
,	O
Facebook	B-System
,	O
Kakoa	B-System
,	O
Viber	B-System
,	O
and	O
Skype	B-System
by	O
making	O
/data/data	O
directories	O
of	O
the	O
apps	O
world	O
readable	O
.	O

Content	O
observers	O
:	O
use	O
Android	B-System
's	O
ContentObserver	O
framework	O
to	O
gather	O
changes	O
in	O
SMS	B-System
,	O
Calendar	B-System
,	O
Contacts	B-System
,	O
Cell	B-System
info	I-System
,	O
Email	B-System
,	O
WhatsApp	B-System
,	O
Facebook	B-System
,	O
Twitter	B-System
,	O
Kakao	B-System
,	O
Viber	B-System
,	O
and	O
Skype	B-System
.	O

Screenshots	O
:	O
captures	O
an	O
image	O
of	O
the	O
current	O
screen	O
via	O
the	O
raw	O
frame	O
buffer	O
.	O

Keylogging	O
:	O
record	O
input	O
events	O
by	O
hooking	O
IPCThreadState	O
:	O
:Transact	O
from	O
/system/lib/libbinder.so	B-Indicator
,	O
and	O
intercepting	O
android	B-Indicator
:	I-Indicator
:parcel	I-Indicator
with	O
the	O
interface	O
com.android.internal.view.IInputContext	B-Indicator
.	O

RoomTap	O
:	O
silently	O
answers	O
a	O
telephone	O
call	O
and	O
stays	O
connected	O
in	O
the	O
background	O
,	O
allowing	O
the	O
caller	O
to	O
hear	O
conversations	O
within	O
the	O
range	O
of	O
the	O
phone	O
's	O
microphone	O
.	O

If	O
the	O
user	O
unlocks	O
their	O
device	O
,	O
they	O
will	O
see	O
a	O
black	O
screen	O
while	O
the	O
app	O
drops	O
the	O
call	O
,	O
resets	O
call	O
settings	O
and	O
prepares	O
for	O
the	O
user	O
to	O
interact	O
with	O
the	O
device	O
normally	O
.	O

Finally	O
,	O
the	O
app	O
can	O
remove	O
itself	O
through	O
three	O
ways	O
:	O
Via	O
a	O
command	O
from	O
the	O
server	O
Autoremove	O
if	O
the	O
device	O
has	O
not	O
been	O
able	O
to	O
check	O
in	O
to	O
the	O
server	O
after	O
60	O
days	O
Via	O
an	O
antidote	O
file	O
.	O

If	O
/sdcard/MemosForNotes	B-Indicator
was	O
present	O
on	O
the	O
device	O
,	O
the	O
Chrysaor	B-Malware
app	O
removes	O
itself	O
from	O
the	O
device	O
.	O

Samples	O
uploaded	O
to	O
VirusTotal	B-Organization
To	O
encourage	O
further	O
research	O
in	O
the	O
security	O
community	O
,	O
we	O
’	O
ve	O
uploaded	O
these	O
sample	O
Chrysaor	B-Malware
apps	O
to	O
Virus	B-Organization
Total	I-Organization
.	O

Package	O
Name	O
SHA256	O
digest	O
SHA1	O
certificate	O
com.network.android	B-Indicator
ade8bef0ac29fa363fc9afd958af0074478aef650adeb0318517b48bd996d5d5	B-Indicator
44f6d1caa257799e57f0ecaf4e2e216178f4cb3d	B-Indicator
com.network.android	B-Indicator
3474625e63d0893fc8f83034e835472d95195254e1e4bdf99153b7c74eb44d86	B-Indicator
516f8f516cc0fd8db53785a48c0a86554f75c3ba	B-Indicator

Additional	O
digests	O
with	O
links	O
to	O
Chrysaor	B-Malware
As	O
a	O
result	O
of	O
our	O
investigation	O
we	O
have	O
identified	O
these	O
additional	O
Chrysaor-related	B-Malware
apps	O
.	O

Package	O
Name	O
SHA256	O
digest	O
SHA1	O
certificate	O
com.network.android	B-Indicator
98ca5f94638768e7b58889bb5df4584bf5b6af56b188da48c10a02648791b30c	B-Indicator
516f8f516cc0fd8db53785a48c0a86554f75c3ba	B-Indicator
com.network.android	B-Indicator
5353212b70aa096d918e4eb6b49eb5ad8f59d9bec02d089e88802c01e707c3a1	B-Indicator

44f6d1caa257799e57f0ecaf4e2e216178f4cb3d	B-Indicator
com.binary.sms.receiver	B-Indicator
9fae5d148b89001555132c896879652fe1ca633d35271db34622248e048c78ae	B-Indicator
7771af1ad3a3d9c0b4d9b55260bb47c2692722cf	B-Indicator
com.android.copy	B-Indicator
e384694d3d17cd88ec3a66c740c6398e07b8ee401320ca61e26bdf96c20485b4	B-Indicator

7771af1ad3a3d9c0b4d9b55260bb47c2692722cf	B-Indicator
com.android.copy	B-Indicator
12e085ab85db887438655feebd249127d813e31df766f8c7b009f9519916e389	B-Indicator
7771af1ad3a3d9c0b4d9b55260bb47c2692722cf	B-Indicator
com.android.copy	B-Indicator
6348104f8ef22eba5ac8ee737b192887629de987badbb1642e347d0dd01420f8	B-Indicator

31a8633c2cd67ae965524d0b2192e9f14d04d016	B-Indicator
FinFisher	B-Malware
exposed	O
:	O
A	O
researcher	O
’	O
s	O
tale	O
of	O
defeating	O
traps	O
,	O
tricks	O
,	O
and	O
complex	O
virtual	O
machines	O
March	O
1	O
,	O
2018	O
Office	B-System
365	I-System
Advanced	I-System
Threat	I-System
Protection	I-System
(	O
Office	B-System
365	I-System
ATP	I-System
)	O
blocked	O
many	O
notable	O
zero-day	O
exploits	O
in	O
2017	O
.	O

In	O
our	O
analysis	O
,	O
one	O
activity	O
group	O
stood	O
out	O
:	O
NEODYMIUM	B-Malware
.	O

This	O
threat	O
actor	O
is	O
remarkable	O
for	O
two	O
reasons	O
:	O
Its	O
access	O
to	O
sophisticated	O
zero-day	O
exploits	O
for	O
Microsoft	B-Organization
and	O
Adobe	B-Organization
software	O
Its	O
use	O
of	O
an	O
advanced	O
piece	O
of	O
government-grade	O
surveillance	O
spyware	O
FinFisher	B-Malware
,	O
also	O
known	O
as	O
FinSpy	B-Malware
and	O
detected	O
by	O
Microsoft	B-Organization
security	O
products	O
as	O
Wingbird	B-Malware
FinFisher	B-Malware
is	O
such	O
a	O
complex	O
piece	O
of	O
malware	O
that	O
,	O
like	O
other	O
researchers	O
,	O
we	O
had	O
to	O
devise	O
special	O
methods	O
to	O
crack	O
it	O
.	O

We	O
needed	O
to	O
do	O
this	O
to	O
understand	O
the	O
techniques	O
FinFisher	B-Malware
uses	O
to	O
compromise	O
and	O
persist	O
on	O
a	O
machine	O
,	O
and	O
to	O
validate	O
the	O
effectiveness	O
of	O
Office	B-System
365	I-System
ATP	I-System
detonation	O
sandbox	O
,	O
Windows	B-System
Defender	I-System
Advanced	I-System
Threat	I-System
Protection	I-System
(	O
Windows	B-System
Defender	I-System
ATP	I-System
)	O
generic	O
detections	O
,	O
and	O
other	O
Microsoft	B-Organization
security	O
solutions	O
.	O

This	O
task	O
proved	O
to	O
be	O
nontrivial	O
.	O

FinFisher	B-Malware
is	O
not	O
afraid	O
of	O
using	O
all	O
kinds	O
of	O
tricks	O
,	O
ranging	O
from	O
junk	O
instructions	O
and	O
“	O
spaghetti	O
code	O
”	O
to	O
multiple	O
layers	O
of	O
virtual	O
machines	O
and	O
several	O
known	O
and	O
lesser-known	O
anti-debug	O
and	O
defensive	O
measures	O
.	O

Security	O
analysts	O
are	O
typically	O
equipped	O
with	O
the	O
tools	O
to	O
defeat	O
a	O
good	O
number	O
of	O
similar	O
tricks	O
during	O
malware	O
investigations	O
.	O

However	O
,	O
FinFisher	B-Malware
is	O
in	O
a	O
different	O
category	O
of	O
malware	O
for	O
the	O
level	O
of	O
its	O
anti-analysis	O
protection	O
.	O

It	O
’	O
s	O
a	O
complicated	O
puzzle	O
that	O
can	O
be	O
solved	O
by	O
skilled	O
reverse	O
engineers	O
only	O
with	O
good	O
amount	O
of	O
time	O
,	O
code	O
,	O
automation	O
,	O
and	O
creativity	O
.	O

The	O
intricate	O
anti-analysis	O
methods	O
reveal	O
how	O
much	O
effort	O
the	O
FinFisher	B-Malware
authors	O
exerted	O
to	O
keep	O
the	O
malware	O
hidden	O
and	O
difficult	O
to	O
analyze	O
.	O

This	O
exercise	O
revealed	O
tons	O
of	O
information	O
about	O
techniques	O
used	O
by	O
FinFisher	B-Malware
that	O
we	O
used	O
to	O
make	O
Office	B-System
365	I-System
ATP	I-System
more	O
resistant	O
to	O
sandbox	O
detection	O
and	O
Windows	B-System
Defender	I-System
ATP	I-System
to	O
catch	O
similar	O
techniques	O
and	O
generic	O
behaviors	O
.	O

Using	O
intelligence	O
from	O
our	O
in-depth	O
investigation	O
,	O
Windows	B-System
Defender	I-System
ATP	I-System
can	O
raise	O
alerts	O
for	O
malicious	O
behavior	O
employed	O
by	O
FinFisher	B-Malware
(	O
such	O
as	O
memory	O
injection	O
in	O
persistence	O
)	O
in	O
different	O
stages	O
of	O
the	O
attack	O
kill	O
chain	O
.	O

Machine	O
learning	O
in	O
Windows	B-System
Defender	I-System
ATP	I-System
further	O
flags	O
suspicious	O
behaviors	O
observed	O
related	O
to	O
the	O
manipulation	O
of	O
legitimate	O
Windows	B-System
binaries	O
.	O

Figure	O
1	O
.	O

Generic	O
Windows	B-System
Defender	I-System
ATP	I-System
detections	O
trigger	O
alerts	O
on	O
FinFisher	B-Malware
behavior	O
While	O
our	O
analysis	O
has	O
allowed	O
us	O
to	O
immediately	O
protect	O
our	O
customers	O
,	O
we	O
’	O
d	O
like	O
to	O
share	O
our	O
insights	O
and	O
add	O
to	O
the	O
growing	O
number	O
of	O
published	O
analyses	O
by	O
other	O
talented	O
researchers	O
(	O
listed	O
below	O
this	O
blog	O
post	O
)	O
.	O

We	O
hope	O
that	O
this	O
blog	O
post	O
helps	O
other	O
researchers	O
to	O
understand	O
and	O
analyze	O
FinFisher	B-Malware
samples	O
and	O
that	O
this	O
industry-wide	O
information-sharing	O
translate	O
to	O
the	O
protection	O
of	O
as	O
many	O
customers	O
as	O
possible	O
.	O

Spaghetti	O
and	O
junk	O
codes	O
make	O
common	O
analyst	O
tools	O
ineffective	O
In	O
analyzing	O
FinFisher	B-Malware
,	O
the	O
first	O
obfuscation	O
problem	O
that	O
requires	O
a	O
solution	O
is	O
the	O
removal	O
of	O
junk	O
instructions	O
and	O
“	O
spaghetti	O
code	O
”	O
,	O
which	O
is	O
a	O
technique	O
that	O
aims	O
to	O
confuse	O
disassembly	O
programs	O
.	O

Spaghetti	O
code	O
makes	O
the	O
program	O
flow	O
hard	O
to	O
read	O
by	O
adding	O
continuous	O
code	O
jumps	O
,	O
hence	O
the	O
name	O
.	O

An	O
example	O
of	O
FinFisher	B-Malware
’	O
s	O
spaghetti	O
code	O
is	O
shown	O
below	O
.	O

Figure	O
2	O
.	O

The	O
spaghetti	O
code	O
in	O
FinFisher	B-Malware
dropper	O
This	O
problem	O
is	O
not	O
novel	O
,	O
and	O
in	O
common	O
situations	O
there	O
are	O
known	O
reversing	O
plugins	O
that	O
may	O
help	O
for	O
this	O
task	O
.	O

In	O
the	O
case	O
of	O
FinFisher	B-Malware
,	O
however	O
,	O
we	O
could	O
not	O
find	O
a	O
good	O
existing	O
interactive	O
disassembler	O
(	O
IDA	O
)	O
plugin	O
that	O
can	O
normalize	O
the	O
code	O
flow	O
.	O

So	O
we	O
decided	O
to	O
write	O
our	O
own	O
plugin	O
code	O
using	O
IDA	O
Python	B-System
.	O

Armed	O
with	O
this	O
code	O
,	O
we	O
removed	O
this	O
first	O
layer	O
of	O
anti-analysis	O
protection	O
.	O

Removing	O
the	O
junk	O
instructions	O
revealed	O
a	O
readable	O
block	O
of	O
code	O
.	O

This	O
code	O
starts	O
by	O
allocating	O
two	O
chunks	O
of	O
memory	O
:	O
a	O
global	O
1	O
MB	O
buffer	O
and	O
one	O
64	O
KB	O
buffer	O
per	O
thread	O
.	O

The	O
big	O
first	O
buffer	O
is	O
used	O
as	O
index	O
for	O
multiple	O
concurrent	O
threads	O
.	O

A	O
big	O
chunk	O
of	O
data	O
is	O
extracted	O
from	O
the	O
portable	O
executable	O
(	O
PE	O
)	O
file	O
itself	O
and	O
decrypted	O
two	O
times	O
using	O
a	O
custom	O
XOR	O
algorithm	O
.	O

We	O
determined	O
that	O
this	O
chunk	O
of	O
data	O
contains	O
an	O
array	O
of	O
opcode	O
instructions	O
ready	O
to	O
be	O
interpreted	O
by	O
a	O
custom	O
virtual	O
machine	O
program	O
(	O
from	O
this	O
point	O
on	O
referenced	O
generically	O
as	O
“	O
VM	O
”	O
)	O
implemented	O
by	O
FinFisher	B-Malware
authors	O
.	O

Figure	O
3	O
.	O

The	O
stages	O
of	O
the	O
FinFisher	B-Malware
multi-layered	O
protection	O
mechanisms	O
Stage	O
0	O
:	O
Dropper	O
with	O
custom	O
virtual	O
machine	O
The	O
main	O
dropper	O
implements	O
the	O
VM	O
dispatcher	O
loop	O
and	O
can	O
use	O
32	O
different	O
opcodes	O
handlers	O
.	O

Th	O
64KB	O
buffer	O
is	O
used	O
as	O
a	O
VM	O
descriptor	O
data	O
structure	O
to	O
store	O
data	O
and	O
the	O
just-in-time	O
(	O
JIT	O
)	O
generated	O
code	O
to	O
run	O
.	O

The	O
VM	O
dispatcher	O
loop	O
routine	O
ends	O
with	O
a	O
JMP	O
to	O
another	O
routine	O
.	O

In	O
total	O
,	O
there	O
are	O
32	O
different	O
routines	O
,	O
each	O
of	O
them	O
implementing	O
a	O
different	O
opcode	O
and	O
some	O
basic	O
functionality	O
that	O
the	O
malware	O
program	O
may	O
execute	O
.	O

Figure	O
4	O
.	O

A	O
snapshot	B-Malware
of	O
the	O
code	O
that	O
processes	O
each	O
VM	O
opcode	O
and	O
the	O
associate	O
interpreter	O
The	O
presence	O
of	O
a	O
VM	O
and	O
virtualized	O
instruction	O
blocks	O
can	O
be	O
described	O
in	O
simpler	O
terms	O
:	O
Essentially	O
,	O
the	O
creators	O
of	O
FinFisher	B-Malware
interposed	O
a	O
layer	O
of	O
dynamic	O
code	O
translation	O
(	O
the	O
virtual	O
machine	O
)	O
that	O
makes	O
analysis	O
using	O
regular	O
tools	O
practically	O
impossible	O
.	O

Static	O
analysis	O
tools	O
like	O
IDA	O
may	O
not	O
be	O
useful	O
in	O
analyzing	O
custom	O
code	O
that	O
is	O
interpreted	O
and	O
executed	O
through	O
a	O
VM	O
and	O
a	O
new	O
set	O
of	O
instructions	O
.	O

On	O
the	O
other	O
hand	O
,	O
dynamic	O
analysis	O
tools	O
(	O
like	O
debuggers	O
or	O
sandbox	O
)	O
face	O
the	O
anti-debug	O
and	O
anti-analysis	O
tricks	O
hidden	O
in	O
the	O
virtualized	O
code	O
itself	O
that	O
detects	O
sandbox	O
environments	O
and	O
alters	O
the	O
behavior	O
of	O
the	O
malware	O
.	O

At	O
this	O
stage	O
,	O
the	O
analysis	O
can	O
only	O
continue	O
by	O
manually	O
investigating	O
the	O
individual	O
code	O
blocks	O
and	O
opcode	O
handlers	O
,	O
which	O
are	O
highly	O
obfuscated	O
(	O
also	O
using	O
spaghetti	O
code	O
)	O
.	O

Reusing	O
our	O
deobfuscation	O
tool	O
and	O
some	O
other	O
tricks	O
,	O
we	O
have	O
been	O
able	O
to	O
reverse	O
and	O
analyze	O
these	O
opcodes	O
and	O
map	O
them	O
to	O
a	O
finite	O
list	O
that	O
can	O
be	O
used	O
later	O
to	O
automate	O
the	O
analysis	O
process	O
with	O
some	O
scripting	O
.	O

The	O
opcode	O
instructions	O
generated	O
by	O
this	O
custom	O
VM	O
are	O
divided	O
into	O
different	O
categories	O
:	O
Logical	O
opcodes	O
,	O
which	O
implement	O
bit-logic	O
operators	O
(	O
OR	O
,	O
AND	O
,	O
NOT	O
,	O
XOR	O
)	O
and	O
mathematical	O
operators	O
Conditional	O
branching	O
opcodes	O
,	O
which	O
implement	O
a	O
code	O
branch	O
based	O
on	O
conditions	O
(	O
equals	O
to	O
JC	O
,	O
JE	O
,	O
JZ	O
,	O
other	O
similar	O
branching	O
opcodes	O
)	O
Load/Store	O
opcodes	O
,	O
which	O
write	O
to	O
or	O
read	O
from	O
particular	O
addresses	O
of	O
the	O
virtual	O
address	O
space	O
of	O
the	O
process	O
Specialized	O
opcodes	O
for	O
various	O
purposes	O
,	O

like	O
execute	O
specialized	O
machine	O
instruction	O
that	O
are	O
not	O
virtualized	O
We	O
are	O
publishing	O
below	O
the	O
(	O
hopefully	O
)	O
complete	O
list	O
of	O
opcodes	O
used	O
by	O
FinFisher	B-Malware
VM	O
that	O
we	O
found	O
during	O
our	O
analysis	O
and	O
integrated	O
into	O
our	O
de-virtualization	O
script	O
:	O
INDEX	O
MNEMONIC	O
DESCRIPTION	O
0x0	O
EXEC	O
Execute	O
machine	O
code	O
0x1	O
JG	O
Jump	O
if	O
greater/Jump	O
if	O
not	O
less	O
or	O
equal	O
0x2	O
WRITE	O
Write	O
a	O
value	O
into	O
the	O
dereferenced	O
internal	O
VM	O
value	O
(	O
treated	O
as	O
a	O
pointer	O
)	O
0x3	O
JNO	O
Jump	O
if	O
not	O
overflow	O
0x4	O
JLE	O
Jump	O

if	O
less	O
or	O
equal	O
(	O
signed	O
)	O
0x5	O
MOV	O
Move	O
the	O
value	O
of	O
a	O
register	O
into	O
the	O
VM	O
descriptor	O
(	O
same	O
as	O
opcode	O
0x1F	O
)	O
0x6	O
JO	O
Jump	O
if	O
overflow	O
0x7	O
PUSH	O
Push	O
the	O
internal	O
VM	O
value	O
to	O
the	O
stack	O
0x8	O
ZERO	O
Reset	O
the	O
internal	O
VM	O
value	O
to	O
0	O
(	O
zero	O
)	O
0x9	O
JP	O
Jump	O
if	O
parity	O
even	O
0xA	O
WRITE	O
Write	O
into	O
an	O
address	O
0xB	O
ADD	O
Add	O
the	O
value	O
of	O
a	O
register	O
to	O
the	O
internal	O
VM	O
value	O
0xC	O
JNS	O
Jump	O
if	O
not	O
signed	O
0xD	O
JL	O
Jump	O
if	O
less	O
(	O
signed	O
)	O
0xE	O

EXEC	O
Execute	O
machine	O
code	O
and	O
branch	O
0xF	O
JBE	O
Jump	O
if	O
below	O
or	O
equal	O
or	O
Jump	O
if	O
not	O
above	O
0x10	O
SHL	O
Shift	O
left	O
the	O
internal	O
value	O
the	O
number	O
of	O
times	O
specified	O
into	O
the	O
opcodes	O
0x11	O
JA	O
Jump	O
if	O
above/Jump	O
if	O
not	O
below	O
or	O
equal	O
0x12	O
MOV	O
Move	O
the	O
internal	O
VM	O
value	O
into	O
a	O
register	O
0x13	O
JZ	O
JMP	O
if	O
zero	O
0x14	O
ADD	O
Add	O
an	O
immediate	O
value	O
to	O
the	O
internal	O
Vm	O
descriptor	O
0x15	O
JB	O
Jump	O
if	O
below	O
(	O
unsigned	O
)	O
0x16	O
JS	O
Jump	O
if	O
signed	O
0x17	O
EXEC	O
Execute	O

machine	O
code	O
(	O
same	O
as	O
opcode	O
0x0	O
)	O
0x18	O
JGE	O
Jump	O
if	O
greater	O
or	O
equal/Jump	O
if	O
not	O
less	O
0x19	O
DEREF	O
Write	O
a	O
register	O
value	O
into	O
a	O
dereferenced	O
pointer	O
0x1A	O
JMP	O
Special	O
obfuscated	O
“	O
Jump	O
if	O
below	O
”	O
opcode	O
0x1B	O
*	O
Resolve	O
a	O
pointer	O
0x1C	O
LOAD	O
Load	O
a	O
value	O
into	O
the	O
internal	O
VM	O
descriptor	O
0x1D	O
JNE	O
Jump	O
if	O
not	O
equal/Jump	O
if	O
not	O
zero	O
0x1E	O
CALL	O
Call	O
an	O
external	O
function	O
or	O
a	O
function	O
located	O
in	O
the	O
dropper	O
0x1F	O
MOV	O

Move	O
the	O
value	O
of	O
a	O
register	O
into	O
the	O
VM	O
descriptor	O
0x20	O
JNB	O
Jump	O
if	O
not	O
below/Jump	O
if	O
above	O
or	O
equal/Jump	O
if	O
not	O
carry	O
0x21	O
JNP	O
Jump	O
if	O
not	O
parity/Jump	O
if	O
parity	O
odd	O
Each	O
virtual	O
instruction	O
is	O
stored	O
in	O
a	O
special	O
data	O
structure	O
that	O
contains	O
all	O
the	O
information	O
needed	O
to	O
be	O
properly	O
read	O
and	O
executed	O
by	O
the	O
VM	O
.	O

This	O
data	O
structure	O
is	O
24	O
bytes	O
and	O
is	O
composed	O
of	O
some	O
fixed	O
fields	O
and	O
a	O
variable	O
portion	O
that	O
depends	O
on	O
the	O
opcode	O
.	O

Before	O
interpreting	O
the	O
opcode	O
,	O
the	O
VM	O
decrypts	O
the	O
opcode	O
’	O
s	O
content	O
(	O
through	O
a	O
simple	O
XOR	O
algorithm	O
)	O
,	O
which	O
it	O
then	O
relocates	O
(	O
if	O
needed	O
)	O
,	O
using	O
the	O
relocation	O
fields	O
.	O

Here	O
is	O
an	O
approximate	O
diagram	O
of	O
the	O
opcode	O
data	O
structure	O
:	O
Figure	O
5	O
.	O

A	O
graphical	O
representation	O
of	O
the	O
data	O
structure	O
used	O
to	O
store	O
each	O
VM	O
opcode	O
The	O
VM	O
handler	O
is	O
completely	O
able	O
to	O
generate	O
different	O
code	O
blocks	O
and	O
deal	O
with	O
relocated	O
code	O
due	O
to	O
address	O
space	O
layout	O
randomization	O
(	O
ASLR	O
)	O
.	O

It	O
is	O
also	O
able	O
to	O
move	O
code	O
execution	O
into	O
different	O
locations	O
if	O
needed	O
.	O

For	O
instance	O
,	O
in	O
the	O
case	O
of	O
the	O
“	O
Execute	O
”	O
opcode	O
(	O
0x17	O
)	O
,	O
the	O
32-bit	O
code	O
to	O
run	O
is	O
stored	O
entirely	O
into	O
the	O
variable	O
section	O
with	O
the	O
value	O
at	O
offset	O
5	O
specifying	O
the	O
number	O
of	O
bytes	O
to	O
be	O
copied	O
and	O
executed	O
.	O

Otherwise	O
,	O
in	O
the	O
case	O
of	O
conditional	O
opcodes	O
,	O
the	O
variable	O
part	O
can	O
contain	O
the	O
next	O
JIT	O
packet	O
ID	O
or	O
the	O
next	O
relative	O
virtual	O
address	O
(	O
RVA	O
)	O
where	O
code	O
execution	O
should	O
continue	O
.	O

Of	O
course	O
,	O
not	O
all	O
the	O
opcodes	O
are	O
can	O
be	O
easily	O
read	O
and	O
understood	O
due	O
to	O
additional	O
steps	O
that	O
the	O
authors	O
have	O
taken	O
to	O
make	O
analysis	O
extremely	O
complicated	O
.	O

For	O
example	O
,	O
this	O
is	O
how	O
opcode	O
0x1A	O
is	O
implemented	O
:	O
The	O
opcode	O
should	O
represent	O
a	O
JB	O
(	O
Jump	O
if	O
below	O
)	O
function	O
,	O
but	O
it	O
’	O
s	O
implemented	O
through	O
set	O
carry	O
(	O
STC	O
)	O
instruction	O
followed	O
by	O
a	O
JMP	O
into	O
the	O
dispatcher	O
code	O
that	O
will	O
verify	O
the	O
carry	O
flag	O
condition	O
set	O
by	O
STC	O
.	O

Figure	O
6	O
.	O

One	O
of	O
the	O
obfuscation	O
tricks	O
included	O
by	O
the	O
malware	O
authors	O
in	O
a	O
VM	O
opcode	O
dispatcher	O
Even	O
armed	O
with	O
the	O
knowledge	O
we	O
have	O
described	O
so	O
far	O
,	O
it	O
still	O
took	O
us	O
many	O
hours	O
to	O
write	O
a	O
full-fledged	O
opcode	O
interpreter	O
that	O
’	O
s	O
able	O
to	O
reconstruct	O
the	O
real	O
code	O
executed	O
by	O
FinFisher	B-Malware
.	O

Stage	O
1	O
:	O
Loader	O
malware	O
keeps	O
sandbox	O
and	O
debuggers	O
away	O
The	O
first	O
stage	O
of	O
FinFisher	B-Malware
running	O
through	O
this	O
complicated	O
virtual	O
machine	O
is	O
a	O
loader	O
malware	O
designed	O
to	O
probe	O
the	O
system	O
and	O
determine	O
whether	O
it	O
’	O
s	O
running	O
in	O
a	O
sandbox	O
environment	O
(	O
typical	O
for	O
cloud-based	O
detonation	O
solution	O
like	O
Office	B-System
365	I-System
ATP	I-System
)	O
.	O

The	O
loader	O
first	O
dynamically	O
rebuilds	O
a	O
simple	O
import	O
address	O
table	O
(	O
IAT	O
)	O
,	O
resolving	O
all	O
the	O
API	O
needed	O
from	O
Kernel32	O
and	O
NtDll	O
libraries	O
.	O

It	O
then	O
continues	O
executing	O
in	O
a	O
spawned	O
new	O
thread	O
that	O
checks	O
if	O
there	O
are	O
additional	O
undesired	O
modules	O
inside	O
its	O
own	O
virtual	O
address	O
space	O
(	O
for	O
example	O
,	O
modules	O
injected	O
by	O
certain	O
security	O
solutions	O
)	O
.	O

It	O
eventually	O
kills	O
all	O
threads	O
that	O
belong	O
to	O
these	O
undesired	O
modules	O
(	O
using	O
ZwQueryInformationThread	O
native	O
API	O
with	O
ThreadQuerySetWin32StartAddress	O
information	O
class	O
)	O
.	O

The	O
first	O
anti-sandbox	O
technique	O
is	O
the	O
loader	O
checking	O
the	O
code	O
segment	O
.	O

If	O
it	O
’	O
s	O
not	O
0x1B	O
(	O
for	O
32-bit	O
systems	O
)	O
or	O
0x23	O
(	O
for	O
32-bit	O
system	O
under	O
Wow64	O
)	O
,	O
the	O
loader	O
exits	O
.	O

Next	O
,	O
the	O
dropper	O
checks	O
its	O
own	O
parent	O
process	O
for	O
indications	O
that	O
it	O
is	O
running	O
in	O
a	O
sandbox	O
setup	O
.	O

It	O
calculates	O
the	O
MD5	O
hash	O
of	O
the	O
lower-case	O
process	O
image	O
name	O
and	O
terminates	O
if	O
one	O
of	O
the	O
following	O
conditions	O
are	O
met	O
:	O
The	O
MD5	O
hash	O
of	O
the	O
parent	O
process	O
image	O
name	O
is	O
either	O
D0C4DBFA1F3962AED583F6FCE666F8BC	B-Indicator
or	O
3CE30F5FED4C67053379518EACFCF879	B-Indicator
The	O
parent	O
process	O
’	O
s	O
full	O
image	O
path	O
is	O
equal	O
to	O
its	O
own	O
process	O
path	O
If	O
these	O
initial	O
checks	O
are	O
passed	O
,	O
the	O
loader	O
builds	O
a	O
complete	O
IAT	O
by	O
reading	O
four	O
imported	O
libraries	O
from	O
disk	O
(	O
ntdll.dll	B-Indicator

,	O
kernel32.dll	B-Indicator
,	O
advapi32.dll	B-Indicator
,	O
and	O
version.dll	B-Indicator
)	O
and	O
remapping	O
them	O
in	O
memory	O
.	O

This	O
technique	O
makes	O
use	O
of	O
debuggers	O
and	O
software	O
breakpoints	O
useless	O
.	O

During	O
this	O
stage	O
,	O
the	O
loader	O
may	O
also	O
call	O
a	O
certain	O
API	O
using	O
native	O
system	O
calls	O
,	O
which	O
is	O
another	O
way	O
to	O
bypass	O
breakpoints	O
on	O
API	O
and	O
security	O
solutions	O
using	O
hooks	O
.	O

Figure	O
7	O
.	O

FinFisher	B-Malware
loader	O
calling	O
native	O
Windows	B-System
API	O
to	O
perform	O
anti-debugging	O
tricks	O
At	O
this	O
point	O
,	O
the	O
fun	O
in	O
analysis	O
is	O
not	O
over	O
.	O

A	O
lot	O
of	O
additional	O
anti-sandbox	O
checks	O
are	O
performed	O
in	O
this	O
exact	O
order	O
:	O
Check	O
that	O
the	O
malware	O
is	O
not	O
executed	O
under	O
the	O
root	O
folder	O
of	O
a	O
drive	O
Check	O
that	O
the	O
malware	O
file	O
is	O
readable	O
from	O
an	O
external	O
source	O
Check	O
that	O
the	O
hash	O
of	O
base	O
path	O
is	O
not	O
3D6D62AF1A7C8053DBC8E110A530C679	B-Indicator
Check	O
that	O
the	O
full	O
malware	O
path	O
contains	O
only	O
human	O
readable	O
characters	O
(	O
“	O
a-z	O
”	O
,	O
“	O
A-Z	O
”	O
,	O
and	O
“	O
0-9	O
”	O
)	O
Check	O
that	O
no	O
node	O
in	O
the	O
full	O
path	O
contains	O
the	O
MD5	O
string	O
of	O
the	O
malware	O

file	O
Fingerprint	O
the	O
system	O
and	O
check	O
the	O
following	O
registry	O
values	O
:	O
HKLM\SOFTWARE\Microsoft\Cryptography\MachineGuid	B-Indicator
should	O
not	O
be	O
“	O
6ba1d002-21ed-4dbe-afb5-08cf8b81ca32	B-Indicator
”	O
HKLM\SOFTWARE\Microsoft\Windows	B-Indicator
NT\CurrentVersion\DigitalProductId	I-Indicator
should	O
not	O
be	O
“	O
55274-649-6478953-23109	B-Indicator
”	O
,	O
“	O
A22-00001	B-Indicator
”	O
,	O
or	O
“	O
47220	B-Indicator
”	O
HARDWARE\Description\System\SystemBiosDate	B-Indicator
should	O
not	O
contain	O
“	O
01/02/03	O
”	O

Check	O
that	O
the	O
mutex	O
WininetStartupMutex0	O
does	O
not	O
already	O
exist	O
Check	O
that	O
no	O
DLL	O
whose	O
base	O
name	O
has	O
hash	O
value	O
of	O
0xC9CEF3E4	B-Indicator
is	O
mapped	O
into	O
the	O
malware	O
address	O
space	O
The	O
hashes	O
in	O
these	O
checks	O
are	O
most	O
likely	O
correspond	O
to	O
sandbox	O
or	O
security	O
products	O
that	O
the	O
FinFisher	B-Malware
authors	O
want	O
to	O
avoid	O
.	O

Next	O
,	O
the	O
loader	O
checks	O
that	O
it	O
’	O
s	O
not	O
running	O
in	O
a	O
virtualized	O
environment	O
(	O
VMWare	B-System
or	O
Hyper-V	B-System
)	O
or	O
under	O
a	O
debugger	O
.	O

For	O
the	O
hardware	O
virtualization	O
check	O
,	O
the	O
loader	O
obtains	O
the	O
hardware	O
device	O
list	O
and	O
checks	O
if	O
the	O
MD5	O
of	O
the	O
vendor	O
ID	O
is	O
equal	O
to	O
a	O
predefined	O
list	O
.	O

In	O
our	O
tests	O
,	O
the	O
malware	O
sample	O
was	O
able	O
to	O
easily	O
detect	O
both	O
VMWare	B-System
and	O
Hyper-V	B-System
environments	O
through	O
the	O
detection	O
of	O
the	O
virtualized	O
peripherals	O
(	O
for	O
example	O
,	O
Vmware	B-Organization
has	O
VEN_15AD	O
as	O
vendor	O
ID	O
,	O
HyperV	O
has	O
VMBus	O
as	O
bus	O
name	O
)	O
.	O

Office	B-System
365	I-System
ATP	I-System
sandbox	O
employs	O
special	O
mechanisms	O
to	O
avoid	O
being	O
detected	O
by	O
similar	O
checks	O
.	O

The	O
loader	O
’	O
s	O
anti-debugger	O
code	O
is	O
based	O
on	O
the	O
following	O
three	O
methods	O
:	O
The	O
first	O
call	O
aims	O
to	O
destroy	O
the	O
debugger	O
connection	O
:	O
NOTE	O
:	O
This	O
call	O
completely	O
stops	O
the	O
execution	O
of	O
WinDbg	O
and	O
other	O
debuggers	O
The	O
second	O
call	O
tries	O
to	O
detect	O
the	O
presence	O
of	O
a	O
debugger	O
:	O
The	O
final	O
call	O
tries	O
to	O
destroy	O
the	O
possibility	O
of	O
adding	O
software	O
breakpoint	O
:	O
Finally	O
,	O
if	O
the	O
loader	O
is	O
happy	O
with	O
all	O
the	O
checks	O
done	O
so	O
far	O
,	O
based	O
on	O
the	O
victim	O
operating	O
system	O
(	O
32	O
or	O
64-bit	O
)	O
it	O
proceeds	O
to	O
decrypt	O
a	O
set	O
of	O
fake	O
bitmap	O
resources	O
(	O
stage	O
2	O

)	O
embedded	O
in	O
the	O
executable	O
and	O
prepares	O
the	O
execution	O
of	O
a	O
new	O
layer	O
of	O
VM	O
decoding	O
.	O

Each	O
bitmap	O
resource	O
is	O
extracted	O
,	O
stripped	O
of	O
the	O
first	O
0x428	O
bytes	O
(	O
BMP	O
headers	O
and	O
garbage	O
data	O
)	O
,	O
and	O
combined	O
into	O
one	O
file	O
.	O

The	O
block	O
is	O
decrypted	O
using	O
a	O
customized	O
algorithm	O
that	O
uses	O
a	O
key	O
derived	O
from	O
the	O
original	O
malware	O
dropper	O
’	O
s	O
TimeDateStamp	O
field	O
multiplied	O
by	O
5	O
.	O

Figure	O
8	O
.	O

The	O
fake	O
bitmap	O
image	O
embedded	O
as	O
resource	O
The	O
32-bit	O
stage	O
2	O
malware	O
uses	O
a	O
customized	O
loading	O
mechanism	O
(	O
i.e.	O
,	O
the	O
PE	O
file	O
has	O
a	O
scrambled	O
IAT	O
and	O
relocation	O
table	O
)	O
and	O
exports	O
only	O
one	O
function	O
.	O

For	O
the	O
64-bit	O
stage	O
2	O
malware	O
,	O
the	O
code	O
execution	O
is	O
transferred	O
from	O
the	O
loader	O
using	O
a	O
well-known	O
technique	O
called	O
Heaven	O
’	O
s	O
Gate	O
.	O

In	O
the	O
next	O
sections	O
,	O
for	O
simplicity	O
,	O
we	O
will	O
continue	O
the	O
analysis	O
only	O
on	O
the	O
64-bit	O
payload	O
.	O

Figure	O
9	O
.	O

Heaven	O
’	O
s	O
gate	O
is	O
still	O
in	O
use	O
in	O
2017	O
Stage	O
2	O
:	O
A	O
second	O
multi-platform	O
virtual	O
machine	O
The	O
64-bit	O
stage	O
2	O
malware	O
implements	O
another	O
loader	O
combined	O
with	O
another	O
virtual	O
machine	O
.	O

The	O
architecture	O
is	O
quite	O
similar	O
to	O
the	O
one	O
described	O
previously	O
,	O
but	O
the	O
opcodes	O
are	O
slightly	O
different	O
.	O

After	O
reversing	O
these	O
opcodes	O
,	O
we	O
were	O
able	O
to	O
update	O
our	O
interpreter	O
script	O
to	O
support	O
both	O
32-bit	O
and	O
64-bit	O
virtual	O
machines	O
used	O
by	O
FinFisher	B-Malware
.	O

INDEX	O
MNEMONIC	O
DESCRIPTION	O
0x0	O
JMP	O
Special	O
obfuscated	O
conditional	O
Jump	O
(	O
always	O
taken	O
or	O
always	O
ignored	O
)	O
0x1	O
JMP	O
Jump	O
to	O
a	O
function	O
(	O
same	O
as	O
opcode	O
0x10	O
)	O
0x2	O
CALL	O
Call	O
to	O
the	O
function	O
pointed	O
by	O
the	O
internal	O
VM	O
value	O
0x3	O
CALL	O
Optimized	O
CALL	O
function	O
(	O
like	O
the	O
0x1E	O
opcode	O
of	O
the	O
32-bit	O
VM	O
)	O
0x4	O
EXEC	O
Execute	O
code	O
and	O
move	O
to	O
the	O
next	O
packet	O
0x5	O
JMP	O
Jump	O
to	O
an	O
internal	O
function	O
0x6	O
NOP	O
No	O
operation	O
,	O
move	O
to	O
the	O

next	O
packet	O
0x7	O
CALL	O
Call	O
an	O
imported	O
API	O
(	O
whose	O
address	O
is	O
stored	O
in	O
the	O
internal	O
VM	O
value	O
)	O
0x8	O
LOAD	O
Load	O
a	O
value	O
into	O
the	O
VM	O
descriptor	O
structure	O
*	O
0x9	O
STORE	O
Store	O
the	O
internal	O
VM	O
value	O
inside	O
a	O
register	O
0xA	O
WRITE	O
Resolve	O
a	O
pointer	O
and	O
store	O
the	O
value	O
of	O
a	O
register	O
in	O
its	O
content	O
0xB	O
READ	O
Move	O
the	O
value	O
pointed	O
by	O
the	O
VM	O
internal	O
value	O
into	O
a	O
register	O
0xC	O
LOAD	O
Load	O
a	O
value	O
into	O
the	O
VM	O
descriptor	O
structure	O
(	O
not	O
optimized	O
)	O
0xD	O
CMP	O
Compare	O
the	O
value	O
pointed	O
by	O
the	O
internal	O
VM	O
descriptor	O

with	O
a	O
register	O
0xE	O
CMP	O
Compare	O
the	O
value	O
pointed	O
by	O
the	O
internal	O
VM	O
descriptor	O
with	O
an	O
immediate	O
value	O
0xF	O
XCHG	O
Exchange	O
the	O
value	O
pointed	O
by	O
the	O
internal	O
VM	O
descriptor	O
with	O
a	O
register	O
0x10	O
SHL	O
Jump	O
to	O
a	O
function	O
(	O
same	O
as	O
opcode	O
0x1	O
)	O
This	O
additional	O
virtual	O
machine	O
performs	O
the	O
same	O
duties	O
as	O
the	O
one	O
already	O
described	O
but	O
in	O
a	O
64-bit	O
environment	O
.	O

It	O
extracts	O
and	O
decrypts	O
the	O
stage	O
3	O
malware	O
,	O
which	O
is	O
stored	O
in	O
encrypted	O
resources	O
such	O
as	O
fake	O
dialog	O
boxes	O
.	O

The	O
extraction	O
method	O
is	O
the	O
same	O
,	O
but	O
the	O
encryption	O
algorithm	O
(	O
also	O
XOR	O
)	O
is	O
much	O
simpler	O
.	O

The	O
new	O
payload	O
is	O
decrypted	O
,	O
remapped	O
,	O
and	O
executed	O
in	O
memory	O
,	O
and	O
represents	O
the	O
installation	O
and	O
persistence	O
stage	O
of	O
the	O
malware	O
.	O

Stage	O
3	O
:	O
Installer	O
that	O
takes	O
DLL	O
side-loading	O
to	O
a	O
new	O
level	O
Stage	O
3	O
represents	O
the	O
setup	O
program	O
for	O
FinFisher	B-Malware
.	O

It	O
is	O
the	O
first	O
plain	O
stage	O
that	O
does	O
not	O
employ	O
a	O
VM	O
or	O
obfuscation	O
.	O

The	O
code	O
supports	O
two	O
different	O
installation	O
methods	O
:	O
setup	O
in	O
a	O
UAC-enforced	B-System
environment	I-System
(	O
with	O
limited	O
privileges	O
)	O
,	O
or	O
an	O
installation	O
with	O
full-administrative	O
privileges	O
enabled	O
(	O
in	O
cases	O
where	O
the	O
malware	O
gains	O
the	O
ability	O
to	O
run	O
with	O
elevated	O
permissions	O
)	O
.	O

We	O
were	O
a	O
bit	O
disappointed	O
that	O
we	O
did	O
not	O
see	O
traces	O
of	O
a	O
true	O
privilege	B-Vulnerability
escalation	I-Vulnerability
exploit	I-Vulnerability
after	O
all	O
this	O
deobfuscation	O
work	O
,	O
but	O
it	O
seems	O
these	O
FinFisher	B-Malware
samples	O
were	O
designed	O
to	O
work	O
just	O
using	O
UAC	O
bypasses	O
.	O

The	O
setup	O
code	O
receives	O
an	O
installation	O
command	O
from	O
the	O
previous	O
stage	O
.	O

In	O
our	O
test	O
,	O
this	O
command	O
was	O
the	O
value	O
3	O
.	O

The	O
malware	O
creates	O
a	O
global	O
event	O
named	O
0x0A7F1FFAB12BB2	B-Indicator
and	O
drops	O
some	O
files	O
under	O
a	O
folder	O
located	O
in	O
C	B-Indicator
:	I-Indicator
\ProgramData	I-Indicator
or	O
in	O
the	O
user	O
application	O
data	O
folder	O
.	O

The	O
name	O
of	O
the	O
folder	O
and	O
the	O
malware	O
configuration	O
are	O
read	O
from	O
a	O
customized	O
configuration	O
file	O
stored	O
in	O
the	O
resource	O
section	O
of	O
the	O
setup	O
program	O
.	O

Here	O
the	O
list	O
of	O
the	O
files	O
potentially	O
dropped	O
during	O
the	O
installation	O
stage	O
:	O
FILE	O
NAME	O
STAGE	O
DESCRIPTION	O
d3d9.dll	B-Indicator
Stage	O
4	O
Malware	O
loader	O
used	O
for	O
UAC	O
environments	O
with	O
limited	O
privileges	O
;	O
also	O
protected	O
by	O
VM	O
obfuscation	O
aepic.dll	B-Indicator
,	O
sspisrv.dll	B-Indicator
,	O
userenv.dll	B-Indicator
Stage	O
4	O
Malware	O
loader	O
used	O
in	O
presence	O
of	O
administrative	O
privileges	O
;	O
executed	O
from	O
(	O
and	O
injected	O
into	O
)	O
a	O
fake	O
service	O
;	O
also	O
protected	O
by	O
VM	O
obfuscation	O
msvcr90.dll	B-Indicator
Stage	O
5	O
Malware	O
payload	O
injected	O
into	O

the	O
explorer.exe	B-Indicator
or	O
winlogon.exe	B-Indicator
process	O
;	O
also	O
protected	O
by	O
VM	O
obfuscation	O
.cab	O
Config	O
Main	O
configuration	O
file	O
;	O
encrypted	O
setup.cab	B-Indicator
Unknown	O
Last	O
section	O
of	O
the	O
setup	O
executable	O
;	O
content	O
still	O
unknown	O
.7z	O
Plugin	O
Malware	O
plugin	O
used	O
to	O
spy	O
the	O
victim	O
network	O
communications	O
wsecedit.rar	B-Indicator
Stage	O
6	O
Main	O
malware	O
executable	O
After	O
writing	O
some	O
of	O
these	O
files	O
,	O
the	O
malware	O
decides	O
which	O
kind	O
of	O
installation	O
to	O
perform	O
based	O
on	O
the	O
current	O
privilege	O
provided	O
by	O
the	O
hosting	O
process	O
(	O
for	O
example	O
,	O
if	O
a	O
Microsoft	B-System
Office	I-System
process	O
was	O
used	O
as	O
exploit	O
vector	O
)	O
:	O
Installation	O
process	O
under	O

UAC	O
When	O
running	O
under	O
a	O
limited	O
UAC	O
account	O
,	O
the	O
installer	O
extracts	O
d3d9.dll	B-Indicator
and	O
creates	O
a	O
persistence	O
key	O
under	O
HKCU\Software\Microsoft\Windows\Run	B-Indicator
.	O

The	O
malware	O
sets	O
a	O
registry	O
value	O
(	O
whose	O
name	O
is	O
read	O
from	O
the	O
configuration	O
file	O
)	O
to	O
“	O
C	B-Indicator
:	I-Indicator
\Windows\system32\rundll32.exe	I-Indicator
c	B-Indicator
:	I-Indicator
\ProgramData\AuditApp\d3d9.dll	I-Indicator
,	I-Indicator
Control_Run	B-Indicator
”	O
.	O

Before	O
doing	O
this	O
,	O
the	O
malware	O
makes	O
a	O
screenshot	O
of	O
the	O
screen	O
and	O
displays	O
it	O
on	O
top	O
of	O
all	O
other	O
windows	B-System
for	O
few	O
seconds	O
.	O

This	O
indicates	O
that	O
the	O
authors	O
are	O
trying	O
to	O
hide	O
some	O
messages	O
showed	O
by	O
the	O
system	O
during	O
the	O
setup	O
process	O
.	O

When	O
loaded	O
with	O
startup	O
command	O
2	O
,	O
the	O
installer	O
can	O
copy	O
the	O
original	O
explorer.exe	B-Indicator
file	I-Indicator
inside	O
its	O
current	O
running	O
directory	O
and	O
rename	O
d3d9.dll	B-Indicator
to	O
uxtheme.dll	B-Indicator
.	O

In	O
this	O
case	O
the	O
persistence	O
is	O
achieved	O
by	O
loading	O
the	O
original	O
explorer.exe	B-Indicator
from	O
its	O
startup	O
location	O
and	O
,	O
using	O
DLL	O
side-loading	O
,	O
passing	O
the	O
execution	O
control	O
to	O
the	O
stage	O
4	O
malware	O
(	O
discussed	O
in	O
next	O
section	O
)	O
.	O

Finally	O
,	O
the	O
malware	O
spawns	O
a	O
thread	O
that	O
has	O
the	O
goal	O
to	O
load	O
,	O
remap	O
,	O
and	O
relocate	O
the	O
stage	O
5	O
malware	O
.	O

In	O
this	O
context	O
,	O
there	O
is	O
indeed	O
no	O
need	O
to	O
execute	O
the	O
stage	O
4	O
malware	O
.	O

The	O
msvcr90.dll	B-Indicator
file	I-Indicator
is	O
opened	O
,	O
read	O
,	O
and	O
decrypted	O
,	O
and	O
the	O
code	O
execution	O
control	O
is	O
transferred	O
to	O
the	O
RunDll	O
exported	O
routine	O
.	O

In	O
the	O
case	O
of	O
32-bit	O
systems	O
,	O
the	O
malware	O
may	O
attempt	O
a	O
known	O
UAC	O
bypass	O
by	O
launching	O
printui.exe	B-Indicator
system	O
process	O
and	O
using	O
token	O
manipulation	O
with	O
NtFilterToken	O
as	O
described	O
in	O
this	O
blog	O
post	O
.	O

Installation	O
process	O
with	O
administrative	O
privilege	O
This	O
installation	O
method	O
is	O
more	O
interesting	O
because	O
it	O
reveals	O
how	O
the	O
malware	O
tries	O
to	O
achieve	O
stealthier	O
persistence	O
on	O
the	O
machine	O
.	O

The	O
method	O
is	O
a	O
well-known	O
trick	O
used	O
by	O
penetration	O
testers	O
that	O
was	O
automated	O
and	O
generalized	O
by	O
FinFisher	B-Malware
The	O
procedure	O
starts	O
by	O
enumerating	O
the	O
KnownDlls	O
object	O
directory	O
and	O
then	O
scanning	O
for	O
section	O
objects	O
of	O
the	O
cached	O
system	O
DLLs	O
.	O

Next	O
,	O
the	O
malware	O
enumerates	O
all	O
.exe	O
programs	O
in	O
the	O
%	O
System	O
%	O
folder	O
and	O
looks	O
for	O
an	O
original	O
signed	O
Windows	B-System
binary	O
that	O
imports	O
from	O
at	O
least	O
one	O
KnownDll	O
and	O
from	O
a	O
library	O
that	O
is	O
not	O
in	O
the	O
KnownDll	O
directory	O
.	O

When	O
a	O
suitable	O
.exe	O
file	O
candidate	O
is	O
found	O
,	O
it	O
is	O
copied	O
into	O
the	O
malware	O
installation	O
folder	O
(	O
for	O
example	O
,	O
C	B-Indicator
:	I-Indicator
\ProgramData	I-Indicator
)	O
.	O

At	O
this	O
point	O
the	O
malware	O
extracts	O
and	O
decrypts	O
a	O
stub	O
DLL	O
from	O
its	O
own	O
resources	O
(	O
ID	O
101	O
)	O
.	O

It	O
then	O
calls	O
a	O
routine	O
that	O
adds	O
a	O
code	O
section	O
to	O
a	O
target	O
module	O
.	O

This	O
section	O
will	O
contain	O
a	O
fake	O
export	O
table	O
mimicking	O
the	O
same	O
export	O
table	O
of	O
the	O
original	O
system	O
DLL	O
chosen	O
.	O

At	O
the	O
time	O
of	O
writing	O
,	O
the	O
dropper	O
supports	O
aepic.dll	B-Indicator
,	O
sspisrv.dll	B-Indicator
,	O
ftllib.dll	B-Indicator
,	O
and	O
userenv.dll	B-Indicator
to	O
host	O
the	O
malicious	O
FinFisher	B-Malware
payload	O
.	O

Finally	O
,	O
a	O
new	O
Windows	B-System
service	O
is	O
created	O
with	O
the	O
service	O
path	O
pointing	O
to	O
the	O
candidate	O
.exe	O
located	O
in	O
this	O
new	O
directory	O
together	O
with	O
the	O
freshly	O
created	O
,	O
benign-looking	O
DLL	O
.	O

In	O
this	O
way	O
,	O
when	O
the	O
service	O
runs	O
during	O
boot	O
,	O
the	O
original	O
Windows	B-System
executable	O
is	O
executed	O
from	O
a	O
different	O
location	O
and	O
it	O
will	O
automatically	O
load	O
and	O
map	O
the	O
malicious	O
DLL	O
inside	O
its	O
address	O
space	O
,	O
instead	O
of	O
using	O
the	O
genuine	O
system	O
library	O
.	O

This	O
routine	O
is	O
a	O
form	O
of	O
generic	O
and	O
variable	O
generator	O
of	O
DLL	O
side-loading	O
combinations	O
.	O

Figure	O
10	O
.	O

Windows	B-System
Defender	I-System
ATP	I-System
timeline	O
can	O
pinpoint	O
the	O
service	O
DLL	O
side-loading	O
trick	O
(	O
in	O
this	O
example	O
,	O
using	O
fltlib.dll	B-Indicator
)	O
.	O

In	O
the	O
past	O
,	O
we	O
have	O
seen	O
other	O
activity	O
groups	O
like	O
LEAD	O
employ	O
a	O
similar	O
attacker	O
technique	O
named	O
“	O
proxy-library	O
”	O
to	O
achieve	O
persistence	O
,	O
but	O
not	O
with	O
this	O
professionalism	O
.	O

The	O
said	O
technique	O
brings	O
the	O
advantage	O
of	O
avoiding	O
auto-start	O
extensibility	O
points	O
(	O
ASEP	O
)	O
scanners	O
and	O
programs	O
that	O
checks	O
for	O
binaries	O
installed	O
as	O
service	O
(	O
for	O
the	O
latter	O
,	O
the	O
service	O
chosen	O
by	O
FinFisher	B-Malware
will	O
show	O
up	O
as	O
a	O
clean	O
Windows	B-System
signed	O
binary	O
)	O
.	O

The	O
malware	O
cleans	O
the	O
system	O
event	O
logs	O
using	O
OpenEventLog/ClearEventLog	O
APIs	O
,	O
and	O
then	O
terminates	O
the	O
setup	O
procedure	O
with	O
a	O
call	O
to	O
StartService	O
to	O
run	O
the	O
stage	O
4	O
malware	O
.	O

Figure	O
11	O
.	O

The	O
DLL	O
side-loaded	O
stage	O
4	O
malware	O
mimicking	O
a	O
real	O
export	O
table	O
to	O
avoid	O
detection	O
Stage	O
4	O
:	O
The	O
memory	O
loader	O
–	O
Fun	O
injection	O
with	O
GDI	O
function	O
hijacking	O
Depending	O
on	O
how	O
stage	O
4	O
was	O
launched	O
,	O
two	O
different	O
things	O
may	O
happen	O
:	O
In	O
the	O
low-integrity	O
case	O
(	O
under	O
UAC	O
)	O
the	O
installer	O
simply	O
injects	O
the	O
stage	O
5	O
malware	O
into	O
the	O
bogus	O
explorer.exe	B-Indicator
process	O
started	O
earlier	O
and	O
terminates	O
In	O
the	O
high-integrity	O
case	O
(	O
with	O
administrative	O
privileges	O
or	O
after	O
UAC	O
bypass	O
)	O
,	O
the	O
code	O
searches	O
for	O
the	O
process	O
hosting	O
the	O
Plug	O
and	O
Play	O
service	O
(	O
usually	O
svchost.exe	B-Indicator

)	O
loaded	O
in	O
memory	O
and	O
injects	O
itself	O
into	O
it	O
For	O
the	O
second	O
scenario	O
,	O
the	O
injection	O
process	O
works	O
like	O
this	O
:	O
The	O
malware	O
opens	O
the	O
target	O
service	O
process	O
.	O

It	O
allocates	O
and	O
fills	O
four	O
chunks	O
of	O
memory	O
inside	O
the	O
service	O
process	O
.	O

One	O
chunk	O
contains	O
the	O
entire	O
malware	O
DLL	O
code	O
(	O
without	O
PE	O
headers	O
)	O
.	O

Another	O
chunk	O
is	O
used	O
to	O
copy	O
a	O
basic	O
Ntdll	O
and	O
Kernel32	O
import	O
address	O
table	O
.	O

Two	O
chunks	O
are	O
filled	O
with	O
an	O
asynchronous	O
procedure	O
call	O
(	O
APC	O
)	O
routine	O
code	O
and	O
a	O
stub	O
.	O

It	O
opens	O
the	O
service	O
thread	O
of	O
the	O
service	O
process	O
and	O
uses	O
the	O
ZwQueueApcThread	B-Indicator
native	O
API	O
to	O
inject	O
an	O
APC	O
.	O

The	O
APC	O
routine	O
creates	O
a	O
thread	O
in	O
the	O
context	O
of	O
the	O
svchost.exe	B-Indicator
process	O
that	O
will	O
map	O
and	O
execute	O
the	O
stage	O
5	O
malware	O
into	O
the	O
winlogon.exe	B-Indicator
process	O
.	O

The	O
injection	O
method	O
used	O
for	O
winlogon.exe	B-Indicator
is	O
also	O
interesting	O
and	O
quite	O
unusual	O
.	O

We	O
believe	O
that	O
this	O
method	O
is	O
engineered	O
to	O
avoid	O
trivial	O
detection	O
of	O
process	O
injection	O
using	O
the	O
well-detected	O
CreateRemoteThread	B-Indicator
or	O
ZwQueueApcThread	B-Indicator
API	O
.	O

The	O
malware	O
takes	O
these	O
steps	O
:	O
Check	O
if	O
the	O
system	O
master	O
boot	O
record	O
(	O
MBR	O
)	O
contains	O
an	O
infection	O
marker	O
(	O
0xD289C989C089	B-Indicator
8-bytes	O
value	O
at	O
offset	O
0x2C	O
)	O
,	O
and	O
,	O
if	O
so	O
,	O
terminate	O
itself	O
Check	O
again	O
if	O
the	O
process	O
is	O
attached	O
to	O
a	O
debugger	O
(	O
using	O
the	O
techniques	O
described	O
previously	O
)	O
Read	O
,	O
decrypt	O
,	O
and	O
map	O
the	O
stage	O
5	O
malware	O
(	O
written	O
in	O
the	O
previous	O
stage	O
in	O
msvcr90.dll	B-Indicator
)	O
Open	O
winlogon.exe	B-Indicator
process	O
Load	O
user32.dll	B-Indicator
system	O
library	O
and	O
read	O
the	O
KernelCallbackTable	B-Indicator

pointer	O
from	O
its	O
own	O
process	O
environment	O
block	O
(	O
PEB	O
)	O
(	O
Note	O
:	O
The	O
KernelCallbackTable	O
points	O
to	O
an	O
array	O
of	O
graphic	O
functions	O
used	O
by	O
Win32	O
kernel	O
subsystem	O
module	O
win32k.sys	B-Indicator
as	O
call-back	O
into	O
user-mode	O
.	O

)	O
Calculate	O
the	O
difference	O
between	O
this	O
pointer	O
and	O
the	O
User32	O
base	O
address	O
.	O

Copy	O
the	O
stage	O
5	O
DLL	O
into	O
winlogon.exe	B-Indicator
Allocate	O
a	O
chunk	O
of	O
memory	O
in	O
winlogon.exe	B-Indicator
process	O
and	O
copy	O
the	O
same	O
APC	O
routine	O
seen	O
previously	O
Read	O
and	O
save	O
the	O
original	O
pointer	O
of	O
the	O
__fnDWORD	O
internal	O
User32	O
routine	O
(	O
located	O
at	O
offset	O
+0x10	O
of	O
the	O
KernelCallbackTable	O
)	O
and	O
replace	O
this	O
pointer	O
with	O
the	O
address	O
of	O
the	O
APC	O
stub	O
routine	O
After	O
this	O
function	O
pointer	O
hijacking	O
,	O
when	O
winlogon.exe	B-Indicator
makes	O
any	O
graphical	O
call	O
(	O
GDI	O
)	O
,	O
the	O
malicious	O
code	O
can	O
execute	O
without	O
using	O
CreateRemoteThread	O
or	O

similar	O
triggers	O
that	O
are	O
easily	O
detectable	O
.	O

After	O
execution	O
it	O
takes	O
care	O
of	O
restoring	O
the	O
original	O
KernelCallbackTable	O
.	O

Stage	O
5	O
:	O
The	O
final	O
loader	O
takes	O
control	O
The	O
stage	O
5	O
malware	O
is	O
needed	O
only	O
to	O
provide	O
one	O
more	O
layer	O
of	O
obfuscation	O
,	O
through	O
the	O
VM	O
,	O
of	O
the	O
final	O
malware	O
payload	O
and	O
to	O
set	O
up	O
a	O
special	O
Structured	O
Exception	O
Hander	O
routine	O
,	O
which	O
is	O
inserted	O
as	O
Wow64PrepareForException	O
in	O
Ntdll	O
.	O

This	O
special	O
exception	O
handler	O
is	O
needed	O
to	O
manage	O
some	O
memory	O
buffers	O
protection	O
and	O
special	O
exceptions	O
that	O
are	O
used	O
to	O
provide	O
more	O
stealthy	O
execution	O
.	O

After	O
the	O
VM	O
code	O
has	O
checked	O
again	O
the	O
user	O
environment	O
,	O
it	O
proceeds	O
to	O
extract	O
and	O
execute	O
the	O
final	O
un-obfuscated	O
payload	O
sample	O
directly	O
into	O
winlogon.exe	B-Indicator
(	O
alternatively	O
,	O
into	O
explorer.exe	B-Indicator
)	O
process	O
.	O

After	O
the	O
payload	O
is	O
extracted	O
,	O
decrypted	O
,	O
and	O
mapped	O
in	O
the	O
process	O
memory	O
,	O
the	O
malware	O
calls	O
the	O
new	O
DLL	O
entry	O
point	O
,	O
and	O
then	O
the	O
RunDll	O
exported	O
function	O
.	O

The	O
latter	O
implements	O
the	O
entire	O
spyware	O
program	O
.	O

Stage	O
6	O
:	O
The	O
payload	O
is	O
a	O
modular	O
spyware	O
framework	O
for	O
further	O
analysis	O
Our	O
journey	O
to	O
deobfuscating	O
FinFisher	B-Malware
has	O
allowed	O
us	O
to	O
uncover	O
the	O
complex	O
anti-analysis	O
techniques	O
used	O
by	O
this	O
malware	O
,	O
as	O
well	O
as	O
to	O
use	O
this	O
intel	O
to	O
protect	O
our	O
customers	O
,	O
which	O
is	O
our	O
top	O
priority	O
.	O

Analysis	O
of	O
the	O
additional	O
spyware	O
modules	O
is	O
future	O
work	O
.	O

It	O
is	O
evident	O
that	O
the	O
ultimate	O
goal	O
of	O
this	O
program	O
is	O
to	O
steal	O
information	O
.	O

The	O
malware	O
architecture	O
is	O
modular	O
,	O
which	O
means	O
that	O
it	O
can	O
execute	O
plugins	O
.	O

The	O
plugins	O
are	O
stored	O
in	O
its	O
resource	O
section	O
and	O
can	O
be	O
protected	O
by	O
the	O
same	O
VM	O
.	O

The	O
sample	O
we	O
analyzed	O
in	O
October	O
,	O
for	O
example	O
,	O
contains	O
a	O
plugin	O
that	O
is	O
able	O
to	O
spy	O
on	O
internet	O
connections	O
,	O
and	O
can	O
even	O
divert	O
some	O
SSL	O
connections	O
and	O
steal	O
data	O
from	O
encrypted	O
traffic	O
.	O

Some	O
FinFisher	B-Malware
variants	O
incorporate	O
an	O
MBR	B-Indicator
rootkit	I-Indicator
,	O
the	O
exact	O
purpose	O
of	O
which	O
is	O
not	O
clear	O
.	O

Quite	O
possibly	O
,	O
this	O
routine	O
targets	O
older	O
platforms	O
like	O
Windows	B-System
7	I-System
and	O
machines	O
not	O
taking	O
advantage	O
of	O
hardware	O
protections	O
like	O
UEFI	O
and	O
SecureBoot	O
,	O
available	O
on	O
Windows	B-System
10	I-System
.	O

Describing	O
this	O
additional	O
piece	O
of	O
code	O
in	O
detail	O
is	O
outside	O
the	O
scope	B-System
of	O
this	O
analysis	O
and	O
may	O
require	O
a	O
new	O
dedicated	O
blog	O
post	O
.	O

Defense	O
against	O
FinFisher	B-Malware
Exposing	O
as	O
much	O
of	O
FinFisher	B-Malware
’	O
s	O
riddles	O
as	O
possible	O
during	O
this	O
painstaking	O
analysis	O
has	O
allowed	O
us	O
to	O
ensure	O
our	O
customers	O
are	O
protected	O
against	O
this	O
advanced	O
piece	O
of	O
malware	O
.	O

Windows	B-System
10	I-System
S	O
devices	O
are	O
naturally	O
protected	O
against	O
FinFisher	B-Malware
and	O
other	O
threats	O
thanks	O
to	O
the	O
strong	O
code	O
integrity	O
policies	O
that	O
don	O
’	O
t	O
allow	O
unknown	O
unsigned	O
binaries	O
to	O
run	O
(	O
thus	O
stopping	O
FinFisher	B-Malware
’	O
s	O
PE	O
installer	O
)	O
or	O
loaded	O
(	O
blocking	O
FinFisher	B-Malware
’	O
s	O
DLL	O
persistence	O
)	O
.	O

On	O
Windows	B-System
10	I-System
,	O
similar	O
code	O
integrity	O
policies	O
can	O
be	O
configured	O
using	O
Windows	B-System
Defender	I-System
Application	I-System
Control	I-System
.	O

Office	B-System
365	I-System
Advanced	I-System
Threat	I-System
Protection	I-System
secures	O
mailboxes	O
from	O
email	O
campaigns	O
that	O
use	O
zero-day	B-Vulnerability
exploits	I-Vulnerability
to	O
deliver	O
threats	O
like	O
FinFisher	B-Malware
.	O

Office	B-System
365	I-System
ATP	I-System
blocks	O
unsafe	O
attachments	O
,	O
malicious	O
links	O
,	O
and	O
linked-to	O
files	O
using	O
time-of-click	O
protection	O
.	O

Using	O
intel	O
from	O
this	O
research	O
,	O
we	O
have	O
made	O
Office	B-System
365	I-System
ATP	I-System
more	O
resistant	O
to	O
FinFisher	B-Malware
’	O
s	O
anti-sandbox	O
checks	O
.	O

Generic	O
detections	O
,	O
advanced	O
behavioral	O
analytics	O
,	O
and	O
machine	O
learning	O
technologies	O
in	O
Windows	B-System
Defender	I-System
Advanced	I-System
Threat	I-System
Protection	I-System
detect	O
FinFisher	B-Malware
’	O
s	O
malicious	O
behavior	O
throughout	O
the	O
attack	O
kill	O
chain	O
and	O
alert	O
SecOps	O
personnel	O
.	O

Windows	B-System
Defender	I-System
ATP	I-System
also	O
integrates	O
with	O
the	O
Windows	B-System
protection	O
stack	O
so	O
that	O
protections	O
from	O
Windows	B-System
Defender	I-System
AV	I-System
and	O
Windows	B-System
Defender	I-System
Exploit	I-System
Guard	I-System
are	O
reported	O
in	O
Windows	B-System
Defender	I-System
ATP	I-System
portal	O
,	O
enabling	O
SecOps	O
personnel	O
to	O
centrally	O
manage	O
security	O
,	O
and	O
as	O
well	O
as	O
promptly	O
investigate	O
and	O
respond	O
to	O
hostile	O
activity	O
in	O
the	O
network	O
.	O

We	O
hope	O
that	O
this	O
writeup	O
of	O
our	O
journey	O
through	O
all	O
the	O
multiple	O
layers	O
of	O
protection	O
,	O
obfuscation	O
,	O
and	O
anti-analysis	O
techniques	O
of	O
FinFisher	B-Malware
will	O
be	O
useful	O
to	O
other	O
researchers	O
studying	O
this	O
malware	O
.	O

We	O
believe	O
that	O
an	O
industry-wide	O
collaboration	O
and	O
information-sharing	O
is	O
important	O
in	O
defending	O
customers	O
against	O
this	O
complex	O
piece	O
of	O
malware	O
.	O

TUESDAY	O
,	O
APRIL	O
9	O
,	O
2019	O
Gustuff	B-Malware
banking	O
botnet	O
targets	O
Australia	O
EXECUTIVE	O
SUMMARY	O
Cisco	B-Organization
Talos	I-Organization
has	O
uncovered	O
a	O
new	O
Android-based	B-System
campaign	O
targeting	O
Australian	O
financial	O
institutions	O
.	O

As	O
the	O
investigation	O
progressed	O
,	O
Talos	B-Organization
came	O
to	O
understand	O
that	O
this	O
campaign	O
was	O
associated	O
with	O
the	O
"	O
ChristinaMorrow	O
''	O
text	O
message	O
spam	O
scam	O
previously	O
spotted	O
in	O
Australia	O
.	O

Although	O
this	O
malware	O
's	O
credential-harvest	O
mechanism	O
is	O
not	O
particularly	O
sophisticated	O
,	O
it	O
does	O
have	O
an	O
advanced	O
self-preservation	O
mechanism	O
.	O

Even	O
though	O
this	O
is	O
not	O
a	O
traditional	O
remote	O
access	O
tool	O
(	O
RAT	O
)	O
,	O
this	O
campaign	O
seems	O
to	O
target	O
mainly	O
private	O
users	O
.	O

Aside	O
from	O
the	O
credential	O
stealing	O
,	O
this	O
malware	O
also	O
includes	O
features	O
like	O
the	O
theft	O
of	O
users	O
'	O
contact	O
list	O
,	O
collecting	O
phone	O
numbers	O
associated	O
names	O
,	O
and	O
files	O
and	O
photos	O
on	O
the	O
device	O
.	O

But	O
that	O
does	O
n't	O
mean	O
companies	O
and	O
organizations	O
are	O
out	O
of	O
the	O
woods	O
.	O

They	O
should	O
still	O
be	O
on	O
the	O
lookout	O
for	O
these	O
kinds	O
of	O
trojans	O
,	O
as	O
the	O
attackers	O
could	O
target	O
corporate	O
accounts	O
that	O
contain	O
large	O
amounts	O
of	O
money	O
.	O

The	O
information	O
collected	O
by	O
the	O
malware	O
and	O
the	O
control	O
over	O
the	O
victim	O
's	O
mobile	O
device	O
allows	O
their	O
operators	O
to	O
perform	O
more	O
complex	O
social	O
engineering	O
attacks	O
.	O

A	O
motivated	O
attacker	O
can	O
use	O
this	O
trojan	O
to	O
harvest	O
usernames	O
and	O
passwords	O
and	O
then	O
reuse	O
them	O
to	O
login	O
into	O
the	O
organization	O
's	O
system	O
where	O
the	O
victim	O
works	O
.	O

This	O
is	O
a	O
good	O
example	O
where	O
two-factor	O
authentication	O
based	O
on	O
SMS	O
would	O
fail	O
since	O
the	O
attacker	O
can	O
read	O
the	O
SMS	O
.	O

Corporations	O
can	O
protect	O
themselves	O
from	O
these	O
side-channel	O
attacks	O
by	O
deploying	O
client-based	O
two-factor	O
authentication	O
,	O
such	O
as	O
Duo	B-System
Security	I-System
.	O

One	O
of	O
the	O
most	O
impressive	O
features	O
of	O
this	O
malware	O
is	O
its	O
resilience	O
.	O

If	O
the	O
command	O
and	O
control	O
(	O
C2	O
)	O
server	O
is	O
taken	O
down	O
,	O
the	O
malicious	O
operator	O
can	O
still	O
recover	O
the	O
malware	O
control	O
by	O
sending	O
SMS	O
messages	O
directly	O
to	O
the	O
infected	O
devices	O
.	O

This	O
makes	O
the	O
taking	O
down	O
and	O
recovery	O
of	O
the	O
network	O
much	O
harder	O
and	O
poses	O
a	O
considerable	O
challenge	O
for	O
defenders	O
.	O

THE	O
CAMPAIGN	O
The	O
malware	O
's	O
primary	O
infection	O
vector	O
is	O
SMS	O
.	O

Just	O
like	O
the	O
old-school	O
mail	O
worms	O
that	O
used	O
the	O
victim	O
's	O
address	B-System
book	I-System
to	O
select	O
the	O
next	O
victims	O
,	O
this	O
banking	O
trojan	O
's	O
activation	O
cycle	O
includes	O
the	O
exfiltration	O
of	O
the	O
victim	O
's	O
address	O
book	O
.	O

The	O
trojan	O
will	O
receive	O
instructions	O
from	O
the	O
C2	O
to	O
spread	O
.	O

Spread	O
command	O
from	O
C2	O
The	O
victim	O
receives	O
the	O
command	O
sendSMSMass	O
.	O

Usually	O
,	O
this	O
message	O
targets	O
four	O
or	O
five	O
people	O
at	O
a	O
time	O
.	O

The	O
body	O
contains	O
a	O
message	O
and	O
URL	O
.	O

Again	O
,	O
the	O
concept	O
is	O
that	O
new	O
victims	O
are	O
more	O
likely	O
to	O
install	O
the	O
malware	O
if	O
the	O
SMS	O
comes	O
from	O
someone	O
they	O
know	O
.	O

When	O
a	O
victim	O
tries	O
to	O
access	O
the	O
URL	O
in	O
the	O
SMS	O
body	O
,	O
the	O
C2	O
will	O
check	O
if	O
the	O
mobile	O
device	O
meets	O
the	O
criteria	O
to	O
receive	O
the	O
malware	O
(	O
see	O
infrastructure	O
section	O
)	O
.	O

If	O
the	O
device	O
does	O
not	O
meet	O
the	O
criteria	O
,	O
it	O
wo	O
n't	O
receive	O
any	O
data	O
,	O
otherwise	O
,	O
it	O
will	O
be	O
redirected	O
to	O
a	O
second	O
server	O
to	O
receive	O
a	O
copy	O
of	O
the	O
malware	O
to	O
install	O
on	O
their	O
device	O
.	O

The	O
domain	O
on	O
this	O
campaign	O
was	O
registered	O
on	O
Jan.	O
19	O
,	O
2019	O
.	O

However	O
,	O
Talos	B-Organization
has	O
identified	O
that	O
was	O
used	O
at	O
least	O
since	O
November	O
2018	O
.	O

During	O
the	O
investigation	O
,	O
Talos	B-Organization
was	O
also	O
able	O
to	O
determine	O
that	O
the	O
same	O
infrastructure	O
has	O
been	O
used	O
to	O
deploy	O
similar	O
campaigns	O
using	O
different	O
versions	O
of	O
the	O
malware	O
.	O

Distribution	O
of	O
victims	O
.	O

Talos	B-Organization
assess	O
with	O
high	O
confidence	O
that	O
this	O
campaign	O
is	O
targeting	O
Australian	O
financial	O
institutions	O
based	O
on	O
several	O
factors	O
.	O

Our	O
Umbrella	O
telemetry	O
shows	O
that	O
the	O
majority	O
of	O
the	O
request	O
comes	O
from	O
Australia	O
and	O
the	O
majority	O
of	O
the	O
phone	O
numbers	O
infected	O
have	O
the	O
international	O
indicative	O
for	O
Australia	O
.	O

Finally	O
,	O
the	O
specific	O
overlays	O
are	O
designed	O
for	O
Australian	O
financial	O
institutions	O
,	O
and	O
Australia	O
is	O
one	O
of	O
the	O
geographic	O
regions	O
that	O
is	O
accepted	O
by	O
the	O
C2	O
.	O

DNS	O
queries	O
distribution	O
over	O
time	O
The	O
campaign	O
does	O
n't	O
seem	O
to	O
be	O
growing	O
at	O
a	O
fast	O
pace	O
.	O

Our	O
data	O
shows	O
,	O
on	O
average	O
,	O
about	O
three	O
requests	O
per	O
hour	O
to	O
the	O
drop	O
host	O
.	O

This	O
request	O
is	O
only	O
made	O
upon	O
installation	O
,	O
but	O
there	O
is	O
no	O
guarantee	O
that	O
it	O
will	O
be	O
installed	O
.	O

This	O
data	O
,	O
when	O
analyzed	O
with	O
the	O
number	O
of	O
commands	O
to	O
send	O
SMSs	O
that	O
Talos	O
received	O
during	O
the	O
investigation	O
,	O
lead	O
us	O
to	O
conclude	O
that	O
the	O
malicious	O
operator	O
is	O
aggressively	O
spreading	O
the	O
malware	O
,	O
but	O
that	O
does	O
n't	O
seem	O
to	O
result	O
in	O
the	O
same	O
number	O
of	O
new	O
infections	O
.	O

Examples	O
of	O
the	O
overlays	O
available	O
to	O
the	O
malware	O
Above	O
,	O
you	O
can	O
see	O
examples	O
of	O
the	O
injections	O
that	O
distributed	O
to	O
the	O
malware	O
as	O
part	O
of	O
this	O
specific	O
campaign	O
.	O

While	O
doing	O
our	O
investigation	O
we	O
were	O
able	O
to	O
identify	O
other	O
malware	O
packages	O
with	O
different	O
names	O
.	O

Some	O
of	O
these	O
might	O
have	O
been	O
used	O
on	O
old	O
campaigns	O
or	O
were	O
already	O
prepared	O
for	O
new	O
campaigns	O
.	O

MALWARE	O
TECHNICAL	O
DETAILS	O
During	O
our	O
investigation	O
,	O
researchers	O
uncovered	O
a	O
malware	O
known	O
as	O
"	O
Gustuff.	B-Malware
''	O
.	O

Given	O
the	O
lack	O
of	O
indicators	O
of	O
compromise	O
,	O
we	O
decided	O
to	O
check	O
to	O
see	O
if	O
this	O
was	O
the	O
same	O
malware	O
we	O
had	O
been	O
researching	O
.	O

Our	O
Threat	O
Intelligence	O
and	O
Interdiction	O
team	O
found	O
the	O
Gustuff	B-Malware
malware	O
being	O
advertised	O
in	O
the	O
Exploit.in	B-Indicator
forum	O
as	O
a	O
botnet	O
for	O
rent	O
.	O

The	O
seller	O
,	O
known	O
as	O
"	O
bestoffer	O
,	O
''	O
was	O
,	O
at	O
some	O
point	O
,	O
expelled	O
from	O
the	O
forum	O
.	O

Gustuff	B-Malware
advertising	O
screenshot	O
The	O
companies	O
advertised	O
in	O
the	O
image	O
above	O
were	O
from	O
Australia	O
,	O
which	O
matches	O
up	O
with	O
the	O
campaign	O
we	O
researched	O
.	O

The	O
screenshots	O
provided	O
by	O
the	O
author	O
align	O
with	O
the	O
advertised	O
features	O
and	O
the	O
features	O
that	O
we	O
discovered	O
while	O
doing	O
our	O
analysis	O
.	O

Admin	O
panel	O
The	O
administration	O
panel	O
shows	O
the	O
application	O
configuration	O
,	O
which	O
matches	O
the	O
commands	O
from	O
the	O
C2	O
.	O

Country	O
selection	O
The	O
administration	O
console	O
screenshots	O
also	O
show	O
the	O
ability	O
to	O
filter	O
the	O
results	O
by	O
country	O
.	O

In	O
this	O
case	O
,	O
"	O
AU	O
''	O
is	I-Organization
the	O
code	O
shown	O
,	O
which	O
is	O
Australia	O
.	O

Based	O
on	O
this	O
information	O
,	O
Talos	B-Organization
assesses	O
with	O
high	O
confidence	O
that	O
the	O
malware	O
is	O
the	O
same	O
and	O
this	O
is	O
,	O
in	O
fact	O
,	O
the	O
Gustuff	B-Malware
malware	O
.	O

Design	O
In	O
the	O
manifest	O
,	O
the	O
malware	O
requests	O
a	O
large	O
number	O
of	O
permissions	O
.	O

However	O
,	O
it	O
does	O
n't	O
request	O
permissions	O
like	O
BIND_ADMIN	O
.	O

To	O
perform	O
some	O
of	O
its	O
activities	O
,	O
the	O
malware	O
does	O
not	O
need	O
high	O
privileges	O
inside	O
the	O
device	O
,	O
as	O
we	O
will	O
explain	O
ahead	O
.	O

Permissions	O
in	O
the	O
manifest	O
This	O
malware	O
is	O
designed	O
to	O
avoid	O
detection	O
and	O
analysis	O
.	O

It	O
has	O
several	O
protections	O
in	O
place	O
,	O
both	O
in	O
the	O
C2	O
and	O
the	O
malware	O
's	O
code	O
.	O

The	O
code	O
is	O
not	O
only	O
obfuscated	O
but	O
also	O
packed	O
.	O

The	O
packer	O
,	O
besides	O
making	O
the	O
static	O
analysis	O
more	O
complex	O
,	O
will	O
break	O
the	O
standard	O
debugger	O
.	O

Manifest	O
activity	O
declaration	O
Class	O
list	O
inside	O
the	O
dex	O
file	O
The	O
main	O
malware	O
classes	O
are	O
packed	O
,	O
to	O
a	O
point	O
where	O
the	O
class	O
defined	O
in	O
the	O
manifest	O
has	O
a	O
handler	O
for	O
the	O
MAIN	O
category	O
that	O
does	O
not	O
exist	O
in	O
the	O
DEX	O
file	O
.	O

Error	O
when	O
trying	O
to	O
debug	O
the	O
malware	O
using	O
the	O
Android	B-System
Studio	I-System
IDE	I-System
.	O

One	O
of	O
the	O
side	O
effects	O
of	O
this	O
packer	O
is	O
the	O
inability	O
of	O
Android	B-System
Studio	I-System
IDE	I-System
to	O
debug	O
the	O
code	O
.	O

This	O
happens	O
because	O
the	O
IDE	O
executes	O
the	O
code	O
from	O
the	O
Android	B-System
debug	I-System
bridge	I-System
(	O
ADB	O
)	O
by	O
calling	O
the	O
activity	O
declared	O
in	O
the	O
manifest	O
by	O
name	O
.	O

Since	O
the	O
class	O
does	O
not	O
exist	O
at	O
startup	O
,	O
the	O
application	O
does	O
not	O
run	O
on	O
the	O
debugger	O
.	O

Although	O
Talos	B-Malware
analyzed	O
the	O
unpacked	O
version	O
of	O
the	O
code	O
,	O
the	O
packer	O
analysis	O
is	O
beyond	O
the	O
scope	O
of	O
this	O
post	O
.	O

Check	O
code	O
for	O
emulators	O
As	O
part	O
of	O
its	O
defense	O
,	O
the	O
malware	O
payload	O
first	O
checks	O
for	O
emulators	O
to	O
prevent	O
analysis	O
on	O
sandboxes	O
.	O

It	O
checks	O
for	O
different	O
kinds	O
of	O
emulators	O
,	O
including	O
QEMU	B-System
,	O
Genymotion	B-System
,	O
BlueStacks	B-System
and	O
Bignox	B-System
.	O

If	O
the	O
malware	O
determines	O
that	O
is	O
not	O
running	O
on	O
an	O
emulator	O
,	O
it	O
then	O
performs	O
additional	O
checks	O
to	O
ensure	O
that	O
it	O
wo	O
n't	O
be	O
detected	O
.	O

Code	O
to	O
check	O
the	O
existence	O
of	O
SafetyNet	O
Google	B-System
API	I-System
It	O
also	O
checks	O
if	O
the	O
Android	B-System
SafetyNet	O
is	O
active	O
and	O
reporting	O
back	O
to	O
the	O
C2	O
.	O

This	O
helps	O
the	O
C2	O
define	O
what	O
actions	O
it	O
can	O
do	O
before	O
being	O
detected	O
on	O
the	O
mobile	O
device	O
.	O

List	O
of	O
anti-virus	O
packages	O
that	O
are	O
checked	O
The	O
payload	O
goes	O
a	O
long	O
way	O
to	O
protect	O
itself	O
and	O
checks	O
for	O
anti-virus	O
software	O
installed	O
on	O
the	O
mobile	O
device	O
.	O

The	O
trojan	O
uses	O
the	O
Android	B-System
Accessibility	I-System
API	O
to	O
intercept	O
all	O
interactions	O
between	O
the	O
user	O
and	O
the	O
mobile	O
device	O
.	O

The	O
Android	B-System
developer	O
documentation	O
describes	O
the	O
accessibility	O
event	O
class	O
as	O
a	O
class	O
that	O
"	O
represents	O
accessibility	O
events	O
that	O
are	O
seen	O
by	O
the	O
system	O
when	O
something	O
notable	O
happens	O
in	O
the	O
user	O
interface	O
.	O

For	O
example	O
,	O
when	O
a	O
button	O
is	O
clicked	O
,	O
a	O
view	O
is	O
focused	O
,	O
etc	O
.	O

''	O
For	O
each	O
interaction	O
,	O
the	O
malware	O
will	O
check	O
if	O
the	O
generator	O
is	O
a	O
package	O
that	O
belongs	O
to	O
the	O
anti-virus	O
list	O
,	O
the	O
malware	O
will	O
abuse	O
another	O
feature	O
of	O
the	O
Accessibility	B-System
API	I-System
.	O

There	O
is	O
a	O
function	O
called	O
"	O
performGlobalAction	O
''	O
with	O
the	O
description	O
below	O
.	O

Android	B-System
documentation	O
describes	O
that	O
function	O
as	O
"	O
a	O
global	O
action	O
.	O

Such	O
an	O
action	O
can	O
be	O
performed	O
at	O
any	O
moment	O
,	O
regardless	O
of	O
the	O
current	O
application	O
or	O
user	O
location	O
in	O
that	O
application	O
.	O

For	O
example	O
,	O
going	O
back	O
,	O
going	O
home	O
,	O
opening	O
recents	O
,	O
etc	O
.	O

''	O
The	O
trojan	O
calls	O
this	O
function	O
with	O
the	O
action	O
GLOBAL_ACTION_BACK	O
,	O
which	O
equals	O
the	O
pressing	O
of	O
the	O
back	O
button	O
on	O
the	O
device	O
,	O
thus	O
canceling	O
the	O
opening	O
of	O
the	O
anti-virus	O
application	O
.	O

The	O
same	O
event	O
interception	O
is	O
used	O
to	O
place	O
the	O
webview	O
overlay	O
when	O
the	O
user	O
tries	O
to	O
access	O
the	O
targeted	O
applications	O
,	O
allowing	O
it	O
to	O
display	O
its	O
overlay	O
,	O
thus	O
intercepting	O
the	O
credentials	O
.	O

The	O
beaconing	O
only	O
starts	O
after	O
the	O
application	O
is	O
installed	O
and	O
removed	O
from	O
the	O
running	O
tasks	O
.	O

Beaconing	O
information	O
The	O
ID	O
is	O
generated	O
for	O
each	O
installation	O
of	O
the	O
malware	O
,	O
while	O
the	O
token	O
remains	O
unique	O
.	O

Some	O
of	O
the	O
checks	O
performed	O
previously	O
are	O
immediately	O
sent	O
to	O
the	O
C2	O
,	O
like	O
the	O
safetyNet	O
,	O
admin	O
and	O
defaultSMSApp	O
.	O

The	O
beaconing	O
is	O
sent	O
to	O
the	O
URL	O
http	B-Indicator
:	I-Indicator
//	I-Indicator
/api/v2/get.php	I-Indicator
with	O
an	O
interval	O
of	O
60	O
seconds	O
.	O

Answer	O
from	O
the	O
C2	O
The	O
C2	O
will	O
check	O
the	O
country	O
field	O
,	O
if	O
it	O
's	O
empty	O
or	O
if	O
the	O
country	O
is	O
not	O
targeted	O
,	O
it	O
will	O
reply	O
with	O
a	O
"	O
Unauthorized	O
''	O
answer	O
.	O

Otherwise	O
,	O
it	O
will	O
return	O
a	O
JSON	O
encoded	O
"	O
OK	O
,	O
''	O
and	O
if	O
that	O
is	O
the	O
case	O
,	O
the	O
command	O
to	O
be	O
executed	O
.	O

List	O
of	O
available	O
commands	O
The	O
command	O
names	O
are	O
self-explanatory	O
.	O

The	O
command	O
will	O
be	O
issued	O
as	O
an	O
answer	O
to	O
the	O
beaconing	O
,	O
and	O
the	O
result	O
will	O
be	O
returned	O
to	O
the	O
URL	O
http	B-Indicator
:	I-Indicator
//	I-Indicator
/api/v2/set_state.php	I-Indicator
Example	O
of	O
the	O
command	O
"	O
changeServer	O
''	O
The	O
commands	O
are	O
issued	O
in	O
a	O
JSON	O
format	O
,	O
and	O
the	O
obfuscation	O
is	O
part	O
of	O
the	O
malware	O
code	O
and	O
not	O
added	O
by	O
the	O
packer	O
.	O

It	O
is	O
a	O
custom	O
obfuscation	O
partly	O
based	O
on	O
base85	B-Indicator
encoding	I-Indicator
,	O
which	O
is	O
in	O
itself	O
unusual	O
,	O
in	O
malware	O
.	O

Base85	B-Indicator
encoding	I-Indicator
is	O
usually	O
used	O
on	O
pdf	O
and	O
postscript	O
documentsThe	O
configuration	O
of	O
the	O
malware	O
is	O
stored	O
in	O
custom	O
preferences	O
files	O
,	O
using	O
the	O
same	O
obfuscation	O
scheme	O
.	O

Activation	O
cycle	O
As	O
we	O
have	O
explained	O
above	O
,	O
the	O
malware	O
has	O
several	O
defence	O
mechanisms	O
.	O

Beside	O
the	O
obfuscation	O
and	O
the	O
environment	O
checks	O
,	O
the	O
malware	O
also	O
has	O
some	O
interesting	O
anti-sandbox	O
mechanisms	O
.	O

After	O
installation	O
,	O
the	O
user	O
needs	O
to	O
run	O
the	O
application	O
.	O

The	O
user	O
needs	O
to	O
press	O
the	O
"	O
close	O
''	O
button	O
to	O
finish	O
the	O
installation	O
.	O

However	O
,	O
this	O
wo	O
n't	O
close	O
the	O
application	O
,	O
it	O
will	O
send	O
it	O
to	O
the	O
background	O
,	O
instead	O
.	O

While	O
the	O
application	O
is	O
in	O
the	O
background	O
,	O
although	O
the	O
service	O
is	O
already	O
running	O
,	O
the	O
beaconing	O
will	O
not	O
start	O
.	O

The	O
beaconing	O
will	O
only	O
start	O
after	O
the	O
application	O
is	O
removed	O
from	O
the	O
background	O
,	O
ultimately	O
stopping	O
it	O
.	O

This	O
will	O
be	O
the	O
trigger	O
for	O
the	O
service	O
to	O
start	O
the	O
beaconing	O
.	O

As	O
mentioned	O
previously	O
,	O
the	O
beaconing	O
is	O
done	O
every	O
60	O
seconds	O
.	O

However	O
,	O
no	O
command	O
is	O
received	O
from	O
the	O
C2	O
until	O
the	O
inactiveTime	O
field	O
(	O
see	O
beaconing	O
information	O
image	O
above	O
)	O
has	O
at	O
least	O
the	O
value	O
of	O
2000000	O
.	O

This	O
time	O
resets	O
every	O
time	O
the	O
user	O
performs	O
some	O
activity	O
.	O

After	O
the	O
checks	O
,	O
the	O
malware	O
becomes	O
active	O
,	O
but	O
first	O
,	O
it	O
goes	O
through	O
seven	O
steps	O
,	O
each	O
one	O
calling	O
a	O
different	O
command	O
:	O
uploadPhoneNumbers	O
:	O
Exfiltrates	O
all	O
phone	O
numbers	O
that	O
are	O
in	O
the	O
contact	O
list	O
.	O

Aside	O
from	O
the	O
natural	O
value	O
of	O
phone	O
numbers	O
associated	O
with	O
the	O
names	O
of	O
their	O
owners	O
.	O

Using	O
the	O
SMS	O
has	O
an	O
initial	O
infection	O
vector	O
is	O
another	O
possibility	O
for	O
the	O
exfiltration	O
.	O

One	O
of	O
the	O
purposes	O
of	O
the	O
exfiltration	O
of	O
the	O
contact	O
list	O
is	O
to	O
use	O
them	O
to	O
attack	O
other	O
victims	O
using	O
SMS	O
as	O
an	O
initial	O
vector	O
.	O

checkApps	O
:	O
Asks	O
the	O
malware	O
to	O
see	O
if	O
the	O
packages	O
sent	O
as	O
parameters	O
are	O
installed	O
.	O

The	O
malware	O
contains	O
a	O
list	O
of	O
209	O
packages	O
hardcoded	O
in	O
its	O
source	O
code	O
.	O

However	O
,	O
the	O
C2	O
can	O
send	O
an	O
updated	O
list	O
.	O

List	O
of	O
packages	O
received	O
from	O
the	O
C2	O
adminNumber	O
:	O
Setup	O
of	O
the	O
admin	O
phone	O
number	O
.	O

In	O
our	O
case	O
,	O
the	O
administrator	O
phone	O
number	O
belongs	O
to	O
a	O
mobile	O
network	O
in	O
Australia	O
.	O

Phone	O
number	O
for	O
administration	O
changeServer	O
:	O
At	O
this	O
point	O
,	O
the	O
malware	O
changes	O
the	O
C2	O
to	O
a	O
new	O
host	O
,	O
even	O
though	O
the	O
API	O
and	O
communication	O
protocol	O
continues	O
to	O
be	O
the	O
same	O
.	O

Change	O
server	O
request	O
The	O
URL	O
's	O
for	O
the	O
new	O
server	O
is	O
obfuscated	O
,	O
preventing	O
easy	O
network	O
identification	O
.	O

changeActivity	O
:	O
This	O
command	O
will	O
set	O
up	O
the	O
webview	O
to	O
overlay	O
any	O
of	O
the	O
target	O
activities	O
.	O

changeActivity	O
command	O
The	O
webview	O
injects	O
are	O
not	O
hosted	O
on	O
the	O
C2	O
,	O
they	O
are	O
hosted	O
on	O
a	O
completely	O
different	O
server	O
.	O

params	O
:	O
This	O
command	O
allows	O
the	O
malicious	O
operator	O
to	O
change	O
configuration	O
parameters	O
in	O
the	O
malware	O
.	O

During	O
this	O
stage	O
of	O
the	O
activation	O
cycle	O
,	O
the	O
malware	O
increases	O
the	O
beaconing	O
time	O
to	O
avoid	O
detection	O
.	O

Command	O
to	O
change	O
the	O
beaconing	O
changeArchive	O
:	O
The	O
final	O
command	O
of	O
the	O
activation	O
cycle	O
is	O
the	O
download	O
of	O
an	O
archive	O
.	O

This	O
archive	O
is	O
stored	O
in	O
the	O
same	O
host	O
has	O
the	O
webviews	O
.	O

The	O
archive	O
is	O
a	O
ZIP	O
containing	O
several	O
files	O
,	O
which	O
is	O
protected	O
with	O
a	O
password	O
.	O

Change	O
archive	O
command	O
After	O
this	O
activation	O
cycle	O
,	O
the	O
malware	O
will	O
start	O
the	O
collection	O
of	O
information	O
activities	O
and	O
dissemination	O
.	O

Malicious	O
activity	O
Once	O
the	O
activation	O
cycle	O
ends	O
,	O
the	O
trojan	O
will	O
start	O
its	O
malicious	O
activities	O
.	O

These	O
activities	O
depend	O
on	O
the	O
device	O
configuration	O
.	O

Depending	O
if	O
the	O
victim	O
has	O
any	O
of	O
the	O
targeted	O
applications	O
,	O
the	O
anti-virus	O
installed	O
or	O
geographic	O
location	O
,	O
the	O
malware	O
can	O
harvest	O
credentials	O
from	O
the	O
targeted	O
applications	O
,	O
exfiltrate	O
all	O
personal	O
information	O
or	O
simply	O
use	O
the	O
victim	O
's	O
device	O
to	O
send	O
SMS	O
to	O
spread	O
the	O
trojan	O
The	O
malware	O
deploys	O
overlaying	O
webviews	O
to	O
trick	O
the	O
user	O
and	O
eventually	O
steal	O
their	O
login	O
credentials	O
.	O

These	O
are	O
adapted	O
to	O
the	O
information	O
the	O
malicious	O
operator	O
wants	O
to	O
retrieve	O
.	O

The	O
first	O
webview	O
overlay	O
is	O
created	O
on	O
step	O
6	O
of	O
the	O
activation	O
cycle	O
.	O

Pin	O
request	O
overlay	O
This	O
overlay	O
asks	O
the	O
user	O
to	O
provide	O
their	O
PIN	O
to	O
unlock	O
the	O
mobile	O
device	O
,	O
which	O
is	O
immediately	O
exfiltrated	O
to	O
the	O
C2	O
.	O

The	O
last	O
step	O
of	O
the	O
activation	O
cycle	O
is	O
the	O
download	O
of	O
a	O
password-protected	O
ZIP	O
file	O
.	O

This	O
file	O
contains	O
all	O
HTML	O
,	O
CSS	O
and	O
PNG	O
files	O
necessary	O
to	O
create	O
overlays	O
.	O

Talos	O
found	O
189	O
logos	O
from	O
banks	O
to	O
cryptocurrency	O
exchanges	O
inside	O
the	O
archive	O
,	O
all	O
of	O
which	O
could	O
be	O
targeted	O
.	O

The	O
archive	O
also	O
contained	O
all	O
the	O
necessary	O
codes	O
to	O
target	O
Australian	O
financial	O
institutions	O
.	O

The	O
overlays	O
are	O
activated	O
by	O
the	O
malicious	O
operator	O
using	O
the	O
command	O
changeActivity	O
,	O
as	O
seen	O
on	O
step	O
5	O
of	O
the	O
activation	O
cycle	O
.	O

In	O
this	O
case	O
,	O
we	O
can	O
see	O
that	O
the	O
HTML	O
code	O
of	O
the	O
overlay	O
is	O
stored	O
in	O
the	O
C2	O
infrastructure	O
.	O

However	O
,	O
since	O
the	O
archive	O
that	O
is	O
downloaded	O
into	O
the	O
device	O
has	O
all	O
the	O
necessary	O
information	O
and	O
the	O
malicious	O
actor	O
has	O
access	O
to	O
the	O
device	O
via	O
SMS	O
,	O
the	O
malicious	O
operator	O
can	O
keep	O
its	O
activity	O
even	O
without	O
the	O
C2	O
infrastructure	O
.	O

Infrastructure	O
The	O
infrastructure	O
supporting	O
this	O
malware	O
is	O
rather	O
complex	O
.	O

It	O
is	O
clear	O
that	O
on	O
all	O
stages	O
there	O
are	O
at	O
least	O
two	O
layers	O
.	O

The	O
infrastructure	O
has	O
several	O
layers	O
,	O
although	O
not	O
being	O
very	O
dynamic	O
,	O
still	O
has	O
several	O
layers	O
each	O
one	O
providing	O
some	O
level	O
of	O
protection	O
.	O

All	O
the	O
IP	O
addresses	O
belong	O
to	O
the	O
same	O
company	O
Hetzner	B-Organization
,	O
an	O
IP-hosting	O
firm	O
in	O
Germany	O
.	O

COVERAGE	O
Cisco	B-Organization
Cloud	B-System
Web	I-System
Security	I-System
(	O
CWS	O
)	O
or	O
Web	B-System
Security	I-System
Appliance	I-System
(	O
WSA	O
)	O
web	O
scanning	O
prevents	O
access	O
to	O
malicious	O
websites	O
and	O
detects	O
malware	O
used	O
in	O
these	O
attacks	O
.	O

Email	O
Security	O
can	O
block	O
malicious	O
emails	O
sent	O
by	O
threat	O
actors	O
as	O
part	O
of	O
their	O
campaign	O
.	O

Network	O
Security	O
appliances	O
such	O
as	O
Next-Generation	B-System
Firewall	I-System
(	O
NGFW	O
)	O
,	O
Next-Generation	B-System
Intrusion	I-System
Prevention	I-System
System	I-System
(	O
NGIPS	O
)	O
,	O
and	O
Meraki	B-System
MX	I-System
can	O
detect	O
malicious	O
activity	O
associated	O
with	O
this	O
threat	O
.	O

AMP	O
Threat	O
Grid	O
helps	O
identify	O
malicious	O
binaries	O
and	O
build	O
protection	O
into	O
all	O
Cisco	B-Organization
Security	O
products	O
.	O

Umbrella	O
,	O
our	O
secure	O
internet	O
gateway	O
(	O
SIG	O
)	O
,	O
blocks	O
users	O
from	O
connecting	O
to	O
malicious	O
domains	O
,	O
IPs	O
,	O
and	O
URLs	O
,	O
whether	O
users	O
are	O
on	O
or	O
off	O
the	O
corporate	O
network	O
.	O

Open	O
Source	O
SNORTⓇ	O
Subscriber	O
Rule	O
Set	O
customers	O
can	O
stay	O
up	O
to	O
date	O
by	O
downloading	O
the	O
latest	O
rule	O
pack	O
available	O
for	O
purchase	O
on	O
Snort.org	O
.	O

INDICATORS	O
OF	O
COMPROMISE	O
(	O
IOCS	O
)	O
Domains	O
Facebook-photos-au.su	O
Homevideo2-12l.ml	B-Indicator
videohosting1-5j.gq	B-Indicator
URLs	O
hxxp	B-Indicator
:	I-Indicator
//88.99.227	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
26/html2/2018/GrafKey/new-inj-135-3-dark.html	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//88.99.227	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
26/html2/arc92/au483x.zip	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//94.130.106	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
117:8080/api/v1/report/records.php	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//88.99.227	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
26/html2/new-inj-135-3-white.html	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//facebook-photos-au	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
su/ChristinaMorrow	I-Indicator
hxxp	B-Indicator
:	I-Indicator
//homevideo2-12l	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ml/mms3/download_3.php	I-Indicator
IP	O
addresses	O
78.46.201.36	B-Indicator
88.99.170.84	B-Indicator
88.99.227.26	B-Indicator
94.130.106.117	B-Indicator
88.99.174.200	B-Indicator
88.99.189.31	B-Indicator
Hash	O
369fcf48c1eb982088c22f86672add10cae967af82613bee6fb8a3669603dc48	B-Indicator
b2d4fcf03c7a8bf135fbd3073bea450e2e6661ad8ef2ab2058a3c04f81fc3f3e	B-Indicator

8f5d5d8419a4832d175a6028c9e7d445f1e99fdc12170db257df79831c69ae4e	B-Indicator
a5ebcdaf5fd10ec9de85d62e48cc97a4e08c699a7ebdeab0351b86ab1370557d	B-Indicator
84578b9b2c3cc1c7bbfcf4038a6c76ae91dfc82eef5e4c6815627eaf6b4ae6f6	B-Indicator

89eecd91dff4bf42bebbf3aa85aa512ddf661d3e9de4c91196c98f4fc325a018	B-Indicator
9edee3f3d539e3ade61ac2956a6900d93ba3b535b6a76b3a9ee81e2251e25c61	B-Indicator
0e48e5dbc3a60910c1460b382d28e087a580f38f57d3f82d4564309346069bd1	B-Indicator
c113cdd2a5e164dcba157fc4e6026495a1cfbcb0b1a8bf3e38e7eddbb316e01f	B-Indicator

1819d2546d9c9580193827c0d2f5aad7e7f2856f7d5e6d40fd739b6cecdb1e9e	B-Indicator
b213c1de737b72f8dd7185186a246277951b651c64812692da0b9fdf1be5bf15	B-Indicator
453e7827e943cdda9121948f3f4a68d6289d09777538f92389ca56f6e6de03f0	B-Indicator
0246dd4acd9f64ff1508131c57a7b29e995e102c74477d5624e1271700ecb0e2	B-Indicator

88034e0eddfdb6297670d28ed810aef87679e9492e9b3e782cc14d9d1a55db84	B-Indicator
e08f08f4fa75609731c6dd597dc55c8f95dbdd5725a6a90a9f80134832a07f2e	B-Indicator
01c5b637f283697350ca361f241416303ab6123da4c6726a6555ac36cb654b5c	B-Indicator
1fb06666befd581019af509951320c7e8535e5b38ad058069f4979e9a21c7e1c	B-Indicator

6bdfb79f813448b7f1b4f4dbe6a45d1938f3039c93ecf80318cedd1090f7e341	B-Indicator
ADDITIONAL	O
INFORMATION	O
Packages	O
monitored	O
pin.secret.access	B-Indicator
com.chase.sig.android	B-Indicator
com.morganstanley.clientmobile.prod	B-Indicator
com.wf.wellsfargomobile	B-Indicator
com.citi.citimobile	B-Indicator
com.konylabs.capitalone	B-Indicator
com.infonow.bofa	B-Indicator
com.htsu.hsbcpersonalbanking	B-Indicator
com.usaa.mobile.android.usaa	B-Indicator

com.schwab.mobile	B-Indicator
com.americanexpress.android.acctsvcs.us	O
com.pnc.ecommerce.mobile	B-Indicator
com.regions.mobbanking	B-Indicator
com.clairmail.fth	B-Indicator
com.grppl.android.shell.BOS	B-Indicator
com.tdbank	B-Indicator
com.huntington.m	B-Indicator
com.citizensbank.androidapp	B-Indicator
com.usbank.mobilebanking	B-Indicator
com.ally.MobileBanking	B-Indicator
com.key.android	B-Indicator
com.unionbank.ecommerce.mobile.android	B-Indicator
com.mfoundry.mb.android.mb_BMOH071025661	B-Indicator

com.bbt.cmol	B-Indicator
com.sovereign.santander	B-Indicator
com.mtb.mbanking.sc.retail.prod	B-Indicator
com.fi9293.godough	B-Indicator
com.commbank.netbank	B-Indicator
org.westpac.bank	B-Indicator
org.stgeorge.bank	B-Indicator
au.com.nab.mobile	B-Indicator
au.com.bankwest.mobile	B-Indicator
au.com.ingdirect.android	B-Indicator
org.banksa.bank	B-Indicator
com.anz.android	B-Indicator
com.anz.android.gomoney	B-Indicator
com.citibank.mobile.au	B-Indicator
org.bom.bank	B-Indicator
com.latuabancaperandroid	B-Indicator

com.comarch.mobile	B-Indicator
com.jpm.sig.android	B-Indicator
com.konylabs.cbplpat	B-Indicator
by.belinvestbank	B-Indicator
no.apps.dnbnor	B-Indicator
com.arkea.phonegap	B-Indicator
com.alseda.bpssberbank	B-Indicator
com.belveb.belvebmobile	B-Indicator
com.finanteq.finance.ca	B-Indicator
pl.eurobank	B-Indicator
pl.eurobank2	B-Indicator
pl.noblebank.mobile	B-Indicator
com.getingroup.mobilebanking	B-Indicator
hr.asseco.android.mtoken.getin	B-Indicator
pl.getinleasing.mobile	B-Indicator
com.icp.ikasa.getinon	B-Indicator

eu.eleader.mobilebanking.pekao	B-Indicator
softax.pekao.powerpay	B-Indicator
softax.pekao.mpos	B-Indicator
dk.jyskebank.mobilbank	B-Indicator
com.starfinanz.smob.android.bwmobilbanking	B-Indicator
eu.newfrontier.iBanking.mobile.SOG.Retail	B-Indicator
com.accessbank.accessbankapp	B-Indicator
com.sbi.SBIFreedomPlus	B-Indicator
com.zenithBank.eazymoney	B-Indicator
net.cts.android.centralbank	B-Indicator
com.f1soft.nmbmobilebanking.activities.main	B-Indicator
com.lb.smartpay	B-Indicator
com.mbmobile	B-Indicator

com.db.mobilebanking	B-Indicator
com.botw.mobilebanking	B-Indicator
com.fg.wallet	B-Indicator
com.sbi.SBISecure	B-Indicator
com.icsfs.safwa	B-Indicator
com.interswitchng.www	B-Indicator
com.dhanlaxmi.dhansmart.mtc	B-Indicator
com.icomvision.bsc.tbc	B-Indicator
hr.asseco.android.jimba.cecro	B-Indicator
com.vanso.gtbankapp	B-Indicator
com.fss.pnbpsp	B-Indicator
com.mfino.sterling	B-Indicator
cy.com.netinfo.netteller.boc	B-Indicator
ge.mobility.basisbank	B-Indicator
com.snapwork.IDBI	B-Indicator

com.lcode.apgvb	B-Indicator
com.fact.jib	B-Indicator
mn.egolomt.bank	B-Indicator
com.pnbrewardz	B-Indicator
com.firstbank.firstmobile	B-Indicator
wit.android.bcpBankingApp.millenniumPL	B-Indicator
com.grppl.android.shell.halifax	B-Indicator
com.revolut.revolut	B-Indicator
de.commerzbanking.mobil	B-Indicator
uk.co.santander.santanderUK	B-Indicator
se.nordea.mobilebank	B-Indicator
com.snapwork.hdfc	B-Indicator
com.csam.icici.bank.imobile	B-Indicator
com.msf.kbank.mobile	B-Indicator

com.bmm.mobilebankingapp	B-Indicator
net.bnpparibas.mescomptes	B-Indicator
fr.banquepopulaire.cyberplus	B-Indicator
com.caisseepargne.android.mobilebanking	B-Indicator
com.palatine.android.mobilebanking.prod	B-Indicator
com.ocito.cdn.activity.creditdunord	B-Indicator
com.fullsix.android.labanquepostale.accountaccess	B-Indicator
mobi.societegenerale.mobile.lappli	B-Indicator
com.db.businessline.cardapp	B-Indicator
com.skh.android.mbanking	B-Indicator
com.ifs.banking.fiid1491	B-Indicator

de.dkb.portalapp	B-Indicator
pl.pkobp.ipkobiznes	B-Indicator
pl.com.suntech.mobileconnect	B-Indicator
eu.eleader.mobilebanking.pekao.firm	B-Indicator
pl.mbank	B-Indicator
pl.upaid.nfcwallet.mbank	B-Indicator
eu.eleader.mobilebanking.bre	B-Indicator
pl.asseco.mpromak.android.app.bre	B-Indicator
pl.asseco.mpromak.android.app.bre.hd	B-Indicator
pl.mbank.mnews	B-Indicator
eu.eleader.mobilebanking.raiffeisen	B-Indicator
pl.raiffeisen.nfc	B-Indicator
hr.asseco.android.jimba.rmb	B-Indicator

com.advantage.RaiffeisenBank	B-Indicator
pl.bzwbk.ibiznes24	B-Indicator
pl.bzwbk.bzwbk24	B-Indicator
pl.bzwbk.mobile.tab.bzwbk24	B-Indicator
com.comarch.mobile.investment	B-Indicator
com.android.vending	B-Indicator
com.snapchat.android	B-Indicator
jp.naver.line.android	B-Indicator
com.viber.voip	B-Indicator
com.gettaxi.android	B-Indicator
com.whatsapp	B-Indicator
com.tencent.mm	B-Indicator
com.skype.raider	B-Indicator
com.ubercab	B-Indicator
com.paypal.android.p2pmobile	B-Indicator

com.circle.android	B-Indicator
com.coinbase.android	B-Indicator
com.walmart.android	B-Indicator
com.bestbuy.android	B-Indicator
com.ebay.gumtree.au	B-Indicator
com.ebay.mobile	B-Indicator
com.westernunion.android.mtapp	B-Indicator
com.moneybookers.skrillpayments	B-Indicator
com.gyft.android	B-Indicator
com.amazon.mShop.android.shopping	B-Indicator
com.comarch.mobile.banking.bgzbnpparibas.biznes	B-Indicator
pl.bnpbgzparibas.firmapp	B-Indicator
com.finanteq.finance.bgz	B-Indicator
pl.upaid.bgzbnpp	B-Indicator

de.postbank.finanzassistent	B-Indicator
pl.bph	B-Indicator
de.comdirect.android	B-Indicator
com.starfinanz.smob.android.sfinanzstatus	B-Indicator
de.sdvrz.ihb.mobile.app	B-Indicator
pl.ing.mojeing	B-Indicator
com.ing.mobile	B-Indicator
pl.ing.ingksiegowosc	B-Indicator
com.comarch.security.mobilebanking	B-Indicator
com.comarch.mobile.investment.ing	B-Indicator
com.ingcb.mobile.cbportal	B-Indicator
de.buhl.finanzblick	B-Indicator
pl.pkobp.iko	B-Indicator
pl.ipko.mobile	B-Indicator
pl.inteligo.mobile	B-Indicator
de.number26.android	B-Indicator

pl.millennium.corpApp	I-Indicator
eu.transfer24.app	O
pl.aliorbank.aib	O
pl.corelogic.mtoken	B-Indicator
alior.bankingapp.android	B-Indicator
com.ferratumbank.mobilebank	B-Indicator
com.swmind.vcc.android.bzwbk_mobile.app	B-Indicator
de.schildbach.wallet	B-Indicator
piuk.blockchain.android	B-Indicator
com.bitcoin.mwallet	B-Indicator
com.btcontract.wallet	B-Indicator
com.bitpay.wallet	B-Indicator
com.bitpay.copay	B-Indicator
btc.org.freewallet.app	B-Indicator
org.electrum.electrum	B-Indicator

com.xapo	B-Indicator
com.airbitz	B-Indicator
com.kibou.bitcoin	B-Indicator
com.qcan.mobile.bitcoin.wallet	B-Indicator
me.cryptopay.android	B-Indicator
com.bitcoin.wallet	B-Indicator
lt.spectrofinance.spectrocoin.android.wallet	B-Indicator
com.kryptokit.jaxx	B-Indicator
com.wirex	B-Indicator
bcn.org.freewallet.app	B-Indicator
com.hashengineering.bitcoincash.wallet	B-Indicator
bcc.org.freewallet.app	B-Indicator
com.coinspace.app	B-Indicator
btg.org.freewallet.app	B-Indicator
net.bither	B-Indicator

co.edgesecure.app	B-Indicator
com.arcbit.arcbit	B-Indicator
distributedlab.wallet	B-Indicator
de.schildbach.wallet_test	B-Indicator
com.aegiswallet	B-Indicator
com.plutus.wallet	B-Indicator
com.coincorner.app.crypt	B-Indicator
eth.org.freewallet.app	B-Indicator
secret.access	B-Indicator
secret.pattern	B-Indicator
RuMMS	B-Malware
:	O
The	O
Latest	O
Family	O
of	O
Android	B-System
Malware	O
Attacking	O
Users	O
in	O
Russia	O
Via	O
SMS	O
Phishing	O
April	O
26	O
,	O
2016	O
Introduction	O
Recently	O
we	O
observed	O
an	O
Android	B-Malware
malware	O
family	O
being	O
used	O
to	O
attack	O
users	O
in	O
Russia	O
.	O

The	O
malware	O
samples	O
were	O
mainly	O
distributed	O
through	O
a	O
series	O
of	O
malicious	O
subdomains	O
registered	O
under	O
a	O
legitimate	O
domain	O
belonging	O
to	O
a	O
well-known	O
shared	O
hosting	O
service	O
provider	O
in	O
Russia	O
.	O

Because	O
all	O
the	O
URLs	O
used	O
in	O
this	O
campaign	O
have	O
the	O
form	O
of	O
hxxp	B-Indicator
:	I-Indicator
//yyyyyyyy	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
XXXX.ru/mms.apk	I-Indicator
(	O
where	O
XXXX.ru	B-Indicator
represents	O
the	O
hosting	O
provider	O
’	O
s	O
domain	O
)	O
,	O
we	O
named	O
this	O
malware	O
family	O
RuMMS	B-Malware
.	O

To	O
lure	O
the	O
victims	O
to	O
download	O
the	O
malware	O
,	O
threat	O
actors	O
use	O
SMS	O
phishing	O
–	O
sending	O
a	O
short	O
SMS	O
message	O
containing	O
a	O
malicious	O
URL	O
to	O
the	O
potential	O
victims	O
.	O

Unwary	O
users	O
who	O
click	O
the	O
seemingly	O
innocuous	O
link	O
will	O
have	O
their	O
device	O
infected	O
with	O
RuMMS	B-Malware
malware	O
.	O

Figure	O
1	O
describes	O
this	O
infection	O
process	O
and	O
the	O
main	O
behaviors	O
of	O
RuMMS	B-Malware
.	O

On	O
April	O
3	O
,	O
2016	O
,	O
we	O
still	O
observed	O
new	O
RuMMS	B-Malware
samples	O
emerging	O
in	O
the	O
wild	O
.	O

The	O
earliest	O
identified	O
sample	O
,	O
however	O
,	O
can	O
be	O
traced	O
back	O
to	O
Jan.	O
18	O
,	O
2016	O
.	O

Within	O
this	O
time	O
period	O
,	O
we	O
identified	O
close	O
to	O
300	O
samples	O
belonging	O
to	O
this	O
family	O
(	O
all	O
sample	O
hashes	O
are	O
listed	O
in	O
the	O
Appendix	O
)	O
.	O

After	O
landing	O
on	O
the	O
victim	O
’	O
s	O
phone	O
,	O
the	O
RuMMS	B-Malware
apps	O
will	O
request	O
device	O
administrator	O
privileges	O
,	O
remove	O
their	O
icons	O
to	O
hide	O
themselves	O
from	O
users	O
,	O
and	O
remain	O
running	O
in	O
the	O
background	O
to	O
perform	O
a	O
series	O
of	O
malicious	O
behaviors	O
.	O

So	O
far	O
we	O
have	O
identified	O
the	O
following	O
behaviors	O
:	O
Sending	O
device	O
information	O
to	O
a	O
remote	O
command	O
and	O
control	O
(	O
C2	O
)	O
server	O
.	O

Contacting	O
the	O
C2	O
server	O
for	O
instructions	O
.	O

Sending	O
SMS	O
messages	O
to	O
financial	O
institutions	O
to	O
query	O
account	O
balances	O
.	O

Uploading	O
any	O
incoming	O
SMS	O
messages	O
(	O
including	O
the	O
balance	O
inquiry	O
results	O
)	O
to	O
the	O
remote	O
C2	O
server	O
.	O

Sending	O
C2-specified	O
SMS	O
messages	O
to	O
phone	O
numbers	O
in	O
the	O
victim	O
’	O
s	O
contacts	O
.	O

Forward	O
incoming	O
phone	O
calls	O
to	O
intercept	O
voice-based	O
two-factor	O
authentication	O
.	O

Each	O
of	O
these	O
behaviors	O
is	O
under	O
the	O
control	O
of	O
the	O
remote	O
C2	O
server	O
.	O

In	O
other	O
words	O
,	O
the	O
C2	O
server	O
can	O
specify	O
the	O
message	O
contents	O
to	O
be	O
sent	O
,	O
the	O
time	O
period	O
in	O
which	O
to	O
forward	O
the	O
voice	O
call	O
,	O
and	O
the	O
recipients	O
of	O
outgoing	O
messages	O
.	O

As	O
part	O
of	O
our	O
investigation	O
into	O
this	O
malware	O
,	O
we	O
emulated	O
an	O
infected	O
Android	B-System
device	O
in	O
order	O
to	O
communicate	O
with	O
the	O
RuMMS	B-Malware
C2	O
server	O
.	O

During	O
one	O
session	O
,	O
the	O
C2	O
server	O
commanded	O
our	O
emulated	O
device	O
to	O
send	O
four	O
different	O
SMS	O
messages	O
to	O
four	O
different	O
phone	O
numbers	O
,	O
all	O
of	O
which	O
were	O
associated	O
with	O
Russian	O
financial	O
institutions	O
.	O

At	O
least	O
three	O
of	O
the	O
messages	O
were	O
intended	O
to	O
check	O
a	O
user	O
’	O
s	O
account	O
balance	O
at	O
the	O
institution	O
(	O
we	O
could	O
not	O
confirm	O
the	O
purpose	O
of	O
the	O
fourth	O
)	O
.Through	O
additional	O
research	O
,	O
we	O
identified	O
several	O
forum	O
posts	O
where	O
victims	O
complained	O
of	O
funds	O
(	O
up	O
to	O
600	O
rubles	O
)	O
were	O
transferred	O
out	O
of	O
their	O
accounts	O
after	O
RuMMS	B-Malware
infected	O
their	O
phones	O
.	O

We	O
do	O
not	O
know	O
exactly	O
how	O
many	O
people	O
have	O
been	O
infected	O
with	O
RuMMS	B-Malware
malware	O
.	O

However	O
,	O
our	O
data	O
suggests	O
that	O
there	O
have	O
been	O
at	O
least	O
2,729	O
infections	O
between	O
January	O
2016	O
and	O
early	O
April	O
2016	O
,	O
with	O
a	O
peak	O
in	O
March	O
of	O
more	O
than	O
1,100	O
infections	O
.	O

Smishing	O
:	O
The	O
Major	O
Way	O
To	O
Distribute	O
RuMMS	B-Malware
We	O
have	O
not	O
observed	O
any	O
instances	O
of	O
RuMMS	B-Malware
on	O
Google	B-System
Play	I-System
or	O
other	O
online	O
app	O
stores	O
.	O

Smishing	O
(	O
SMS	O
phishing	O
)	O
is	O
currently	O
the	O
primary	O
way	O
threat	O
actors	O
are	O
distributing	O
the	O
malware	O
.	O

The	O
process	O
starts	O
when	O
an	O
SMS	O
phishing	O
message	O
arrives	O
at	O
a	O
user	O
’	O
s	O
phone	O
.	O

An	O
example	O
SMS	O
message	O
is	O
shown	O
in	O
Figure	O
1	O
.	O

The	O
message	O
translates	O
roughly	O
to	O
“	O
You	O
got	O
a	O
photo	O
in	O
MMS	O
format	O
:	O
hxxp	B-Indicator
:	I-Indicator
//yyyyyyyy.XXXX.ru/mms.apk.	I-Indicator
”	O
So	O
far	O
we	O
identified	O
seven	O
different	O
URLs	O
being	O
used	O
to	O
spread	O
RuMMS	B-Malware
in	O
the	O
wild	O
.	O

All	O
of	O
the	O
URLs	O
reference	O
the	O
file	O
“	O
mms.apk	B-Indicator
”	O
and	O
all	O
use	O
the	O
domain	O
“	O
XXXX.ru	B-Indicator
”	O
,	O
which	O
belongs	O
to	O
a	O
top	O
five	O
shared	O
hosting	O
platform	O
in	O
Russia	O
(	O
the	O
domain	O
itself	O
has	O
been	O
obfuscated	O
to	O
anonymize	O
the	O
provider	O
)	O
.	O

The	O
threat	O
actors	O
registered	O
at	O
least	O
seven	O
subdomains	O
through	O
the	O
hosting	O
provider	O
,	O
each	O
consisting	O
of	O
eight	O
random-looking	O
characters	O
(	O
asdfgjcr	B-Indicator
,	O
cacama18	B-Indicator
,	O
cacamadf	B-Indicator
,	O
konkonq2	B-Indicator
,	O
mmsmtsh5	B-Indicator
,	O
riveroer	B-Indicator
,	O
and	O
sdfkjhl2	B-Indicator
.	O

)	O
As	O
of	O
this	O
writing	O
,	O
no	O
files	O
were	O
hosted	O
at	O
any	O
of	O
the	O
links	O
.	O

The	O
threat	O
actors	O
seem	O
to	O
have	O
abandoned	O
these	O
URLs	O
and	O
might	O
be	O
looking	O
into	O
other	O
ways	O
to	O
reach	O
more	O
victims	O
.	O

Use	O
of	O
a	O
shared	O
hosting	O
service	O
to	O
distribute	O
malware	O
is	O
highly	O
flexible	O
and	O
low	O
cost	O
for	O
the	O
threat	O
actors	O
.	O

It	O
is	O
also	O
much	O
harder	O
for	O
network	O
defenders	O
or	O
researchers	O
to	O
track	O
a	O
campaign	O
where	O
the	O
infrastructure	O
is	O
a	O
moving	O
target	O
.	O

Many	O
top	O
providers	O
in	O
Russia	O
offer	O
cheap	O
prices	O
for	O
their	O
shared	O
hosting	O
services	O
,	O
and	O
some	O
even	O
provide	O
free	O
30-day	O
trial	O
periods	O
.	O

Threat	O
actors	O
can	O
register	O
subdomains	O
through	O
the	O
hosting	O
provider	O
and	O
use	O
the	O
provider	O
’	O
s	O
services	O
for	O
a	O
short-period	O
campaign	O
.	O

A	O
few	O
days	O
later	O
they	O
can	O
cancel	O
the	O
trial	O
and	O
do	O
not	O
need	O
to	O
pay	O
a	O
penny	O
.	O

In	O
addition	O
,	O
these	O
out-of-the-box	O
hosting	O
services	O
usually	O
provide	O
better	O
infrastructure	O
than	O
the	O
attackers	O
could	O
manage	O
to	O
construct	O
(	O
or	O
compromise	O
)	O
themselves	O
.	O

RuMMS	B-Malware
Code	O
Analysis	O
All	O
RuMMS	B-Malware
samples	O
share	O
the	O
same	O
behaviors	O
,	O
major	O
parts	O
of	O
which	O
are	O
shown	O
in	O
Figure	O
1	O
.	O

However	O
,	O
the	O
underlying	O
code	O
can	O
be	O
quite	O
different	O
in	O
that	O
various	O
obfuscation	O
mechanisms	O
were	O
adopted	O
to	O
evade	O
detection	O
by	O
anti-virus	O
tools	O
.	O

We	O
used	O
a	O
sample	O
app	O
named	O
“	O
org.starsizew	B-Indicator
”	O
with	O
an	O
MD5	O
of	O
d8caad151e07025fdbf5f3c26e3ceaff	B-Indicator
to	O
analyze	O
RuMMS	B-Malware
’	O
s	O
code	O
.	O

Several	O
of	O
the	O
main	O
components	O
of	O
RuMMS	B-Malware
are	O
shown	O
in	O
Figure	O
2	O
.	O

The	O
activity	O
class	O
“	O
org.starsizew.MainActivity	B-Indicator
”	O
executes	O
when	O
the	O
app	O
is	O
started	O
.	O

It	O
first	O
starts	O
another	O
activity	O
defined	O
in	O
“	O
org.starsizew.Aa	B-Indicator
”	O
to	O
request	O
device	O
administrator	O
privileges	O
,	O
and	O
then	O
calls	O
the	O
following	O
API	O
of	O
“	O
android.content.pm.PackageManager	B-Indicator
”	O
(	O
the	O
Android	B-System
package	O
manager	O
to	O
remove	O
its	O
own	O
icon	O
on	O
the	O
home	O
screen	O
in	O
order	O
to	O
conceal	O
the	O
existence	O
of	O
RuMMS	B-Malware
from	O
the	O
user	O
:	O
At	O
the	O
same	O
time	O
,	O
”	O
org.starsizew.MainActivity	B-Indicator
”	O
will	O
start	O
the	O
main	O
service	O
as	O
defined	O
in	O
“	O
org.starsizew.Tb	B-Indicator
”	O
,	O
and	O
use	O
a	O
few	O
mechanisms	O
to	O
keep	O
the	O
main	O
service	O
running	O
continuously	O

in	O
the	O
background	O
.	O

The	O
class	O
“	O
org.starsizew.Ac	B-Indicator
”	O
is	O
designed	O
for	O
this	O
purpose	O
;	O
its	O
only	O
task	O
is	O
to	O
check	O
if	O
the	O
main	O
service	O
is	O
running	O
,	O
and	O
restart	O
the	O
main	O
service	O
if	O
the	O
answer	O
is	O
no	O
.	O

The	O
class	O
“	O
org.starsizew.Tb	B-Indicator
”	O
also	O
has	O
a	O
self-monitoring	O
mechanism	O
to	O
restart	O
itself	O
when	O
its	O
own	O
onDestroy	O
API	O
is	O
triggered	O
.	O

Other	O
than	O
that	O
,	O
its	O
major	O
functionality	O
is	O
to	O
collect	O
private	O
device	O
information	O
,	O
upload	O
it	O
to	O
a	O
remote	O
C2	O
server	O
,	O
and	O
handle	O
any	O
commands	O
as	O
requested	O
by	O
the	O
C2	O
server	O
.	O

All	O
those	O
functions	O
are	O
implemented	O
in	O
asynchronous	O
tasks	O
by	O
“	O
org.starsizew.i	B-Indicator
”	O
.	O

The	O
class	O
“	O
org.starsizew.Ma	B-Indicator
”	O
is	O
registered	O
to	O
intercept	O
incoming	O
SMS	O
messages	O
,	O
the	O
arrival	O
of	O
which	O
will	O
trigger	O
the	O
Android	B-System
system	O
to	O
call	O
its	O
“	O
onReceive	O
”	O
API	O
.	O

Its	O
major	O
functionality	O
is	O
also	O
implemented	O
through	O
the	O
call	O
of	O
the	O
asynchronous	O
task	O
(	O
“	O
org.starsizew.i	B-Indicator
”	O
)	O
,	O
including	O
uploading	O
the	O
incoming	O
SMS	O
messages	O
to	O
the	O
remote	O
C2	O
server	O
and	O
executing	O
any	O
commands	O
as	O
instructed	O
by	O
the	O
remote	O
attacker	O
.	O

C2	O
Communication	O
The	O
C2	O
communication	O
includes	O
two	O
parts	O
:	O
sending	O
information	O
to	O
the	O
remote	O
HTTP	O
server	O
and	O
parsing	O
the	O
server	O
’	O
s	O
response	O
to	O
execute	O
any	O
commands	O
as	O
instructed	O
by	O
the	O
remote	O
attackers	O
.	O

The	O
functionality	O
for	O
these	O
two	O
parts	O
is	O
implemented	O
by	O
doInBackground	O
and	O
onPostExecute	O
respectively	O
,	O
two	O
API	O
methods	O
of	O
“	O
android.os.AsyncTask	B-Indicator
”	O
as	O
extended	O
by	O
class	O
“	O
org.starsizew.i	B-Indicator
”	O
.	O

Figure	O
3	O
.	O

Method	O
doInBackground	O
:	O
to	O
send	O
information	O
to	O
remote	O
C2	O
server	O
As	O
seen	O
from	O
the	O
major	O
code	O
body	O
of	O
method	O
doInBackground	O
shown	O
in	O
Figure	O
3	O
(	O
some	O
of	O
the	O
original	O
classes	O
and	O
methods	O
are	O
renamed	O
for	O
easier	O
understanding	O
)	O
,	O
there	O
are	O
three	O
calls	O
to	O
HttpPost	O
with	O
different	O
contents	O
as	O
parameters	O
.	O

At	O
line	O
5	O
,	O
local	O
variable	O
v4	O
specifies	O
the	O
first	O
parameter	O
url	O
,	O
which	O
can	O
be	O
changed	O
by	O
the	O
remote	O
C2	O
server	O
later	O
.	O

These	O
URLs	O
are	O
all	O
in	O
the	O
form	O
of	O
“	O
http	B-Indicator
:	I-Indicator
//	I-Indicator
$	I-Indicator
C2.	I-Indicator
$	I-Indicator
SERVER.	I-Indicator
$	I-Indicator
IP/api/	I-Indicator
?	I-Indicator

id=	I-Indicator
$	I-Indicator
NUM	I-Indicator
”	O
.	O

The	O
second	O
parameter	O
is	O
a	O
constant	O
string	O
“	O
POST	O
”	O
,	O
and	O
the	O
third	O
parameter	O
is	O
a	O
series	O
of	O
key-value	O
pairs	O
to	O
be	O
sent	O
,	O
assembled	O
at	O
runtime	O
.	O

The	O
value	O
of	O
the	O
first	O
item	O
,	O
whose	O
key	O
is	O
“	O
method	O
”	O
(	O
line	O
7	O
)	O
,	O
indicates	O
the	O
type	O
of	O
the	O
contents	O
:	O
install	O
,	O
info	O
and	O
sms	O
.	O

The	O
first	O
type	O
of	O
content	O
,	O
starting	O
with	O
“	O
method=install	O
”	O
,	O
will	O
be	O
sent	O
when	O
the	O
app	O
is	O
started	O
for	O
the	O
first	O
time	O
,	O
including	O
the	O
following	O
device	O
private	O
information	O
:	O
Victim	O
identifier	O
Network	O
operator	O
Device	O
model	O
Device	O
OS	O
version	O
Phone	O
number	O
Device	O
identifier	O
App	O
version	O
Country	O
The	O
second	O
type	O
of	O
information	O
will	O
be	O
sent	O
periodically	O
to	O
indicate	O
that	O
the	O
device	O
is	O
alive	O
.	O

It	O
only	O
has	O
two	O
parts	O
,	O
the	O
method	O
indicated	O
by	O
word	O
“	O
info	O
”	O
and	O
the	O
victim	O
identifier	O
.	O

The	O
third	O
type	O
of	O
information	O
will	O
be	O
sent	O
when	O
RuMMS	B-Malware
intercepts	O
any	O
SMS	O
messages	O
,	O
including	O
the	O
balance	O
inquiry	O
results	O
when	O
it	O
contacts	O
the	O
SMS	O
code	O
of	O
a	O
particular	O
financial	O
service	O
.	O

Method	O
onPostExecute	O
parses	O
the	O
response	O
from	O
the	O
above	O
HTTP	O
session	O
and	O
executes	O
the	O
commands	O
provided	O
by	O
the	O
remote	O
attacker	O
.	O

As	O
seen	O
from	O
the	O
code	O
in	O
Figure	O
5	O
,	O
the	O
commands	O
RuMMS	B-Malware
supports	O
right	O
now	O
include	O
:	O
install_true	O
:	O
to	O
modify	O
app	O
preference	O
to	O
indicate	O
that	O
the	O
C2	O
server	O
received	O
the	O
victim	O
device	O
’	O
s	O
status	O
.	O

sms_send	O
:	O
to	O
send	O
C2-specified	O
SMS	O
messages	O
to	O
C2-specified	O
recipients	O
.	O

sms_grab	O
:	O
to	O
upload	O
periodically	O
the	O
SMS	O
messages	O
in	O
the	O
inbox	O
to	O
C2	O
server	O
.	O

delivery	O
:	O
to	O
deliver	O
specified	O
text	O
to	O
all	O
victim	O
’	O
s	O
contacts	O
(	O
SMS	O
worming	O
)	O
.	O

call_number	O
:	O
to	O
forward	O
phone	O
calls	O
to	O
intercept	O
voice	O
based	O
two-factor	O
authentication	O
.	O

new_url	O
:	O
to	O
change	O
the	O
URL	O
of	O
the	O
C2	O
server	O
in	O
the	O
app	O
preference	O
.	O

ussd	O
:	O
to	O
call	O
a	O
C2-specified	O
phone	O
number	O
.	O

Figure	O
5	O
.	O

Method	O
onPostExecute	O
:	O
to	O
handle	O
instructions	O
from	O
remote	O
C2	O
Figure	O
6	O
shows	O
an	O
example	O
response	O
sent	O
back	O
from	O
one	O
C2	O
server	O
.	O

Note	O
that	O
inside	O
this	O
single	O
response	O
,	O
there	O
is	O
one	O
“	O
install_true	O
”	O
command	O
,	O
one	O
“	O
sms_grab	O
”	O
command	O
and	O
four	O
“	O
sms_send	O
”	O
commands	O
.	O

With	O
the	O
four	O
“	O
sms_send	O
”	O
commands	O
,	O
the	O
messages	O
as	O
specified	O
in	O
the	O
key	O
“	O
text	O
”	O
will	O
be	O
sent	O
immediately	O
to	O
the	O
specified	O
short	O
numbers	O
.	O

Our	O
analysis	O
suggests	O
that	O
the	O
four	O
short	O
numbers	O
are	O
associated	O
with	O
Russian	O
financial	O
institutions	O
,	O
presumably	O
where	O
a	O
victim	O
would	O
be	O
likely	O
to	O
have	O
accounts	O
.	O

Figure	O
6	O
.	O

Example	O
Response	O
in	O
JSON	O
format	O
In	O
particular	O
,	O
short	O
number	O
“	O
+7494	O
”	O
is	O
associated	O
with	O
a	O
payment	O
service	O
provider	O
in	O
Russia	O
.	O

The	O
provider	O
’	O
s	O
website	O
described	O
how	O
the	O
code	O
7494	O
can	O
be	O
used	O
to	O
provide	O
a	O
series	O
of	O
payment-related	O
capabilities	O
.	O

For	O
example	O
,	O
sending	O
text	O
“	O
Balance	O
”	O
will	O
trigger	O
a	O
response	O
with	O
the	O
victim	O
’	O
s	O
wallet	O
balance	O
.	O

Sending	O
text	O
“	O
confirm	O
1	O
”	O
will	O
include	O
proof	O
of	O
payment	O
.	O

Sending	O
text	O
“	O
call	O
on	O
”	O
will	O
activate	O
the	O
USSD	O
payment	O
confirmation	O
service	O
.	O

During	O
our	O
investigation	O
,	O
we	O
observed	O
the	O
C2	O
server	O
sending	O
multiple	O
“	O
balance	O
”	O
commands	O
to	O
different	O
institutions	O
,	O
presumably	O
to	O
query	O
the	O
victim	O
’	O
s	O
financial	O
account	O
balances	O
.	O

RuMMS	B-Malware
can	O
upload	O
responses	O
to	O
the	O
balance	O
inquiries	O
(	O
received	O
via	O
SMS	O
message	O
)	O
to	O
the	O
remote	O
C2	O
server	O
,	O
which	O
can	O
send	O
back	O
additional	O
commands	O
to	O
be	O
sent	O
from	O
the	O
victim	O
to	O
the	O
provider	O
’	O
s	O
payment	O
service	O
.	O

These	O
could	O
include	O
resetting	O
the	O
user	O
’	O
s	O
PIN	O
,	O
enabling	O
or	O
disabling	O
various	O
alerts	O
and	O
confirmations	O
,	O
and	O
confirming	O
the	O
user	O
’	O
s	O
identity	O
.	O

RuMMS	B-Malware
Samples	O
,	O
C2	O
,	O
Hosting	O
Sites	O
,	O
Infections	O
and	O
Timeline	O
In	O
total	O
we	O
captured	O
297	O
RuMMS	B-Malware
samples	O
,	O
all	O
of	O
which	O
attempt	O
to	O
contact	O
an	O
initial	O
C2	O
server	O
that	O
we	O
extracted	O
from	O
the	O
app	O
package	O
.	O

Figure	O
7	O
lists	O
the	O
IP	O
addresses	O
of	O
these	O
C2	O
servers	O
,	O
the	O
number	O
of	O
RuMMS	B-Malware
apps	O
that	O
connect	O
to	O
each	O
of	O
them	O
,	O
and	O
the	O
example	O
URL	O
used	O
as	O
the	O
first	O
parameter	O
of	O
the	O
HttpPost	O
operation	O
(	O
used	O
in	O
the	O
code	O
of	O
Figure	O
3	O
)	O
.	O

This	O
indicates	O
that	O
multiple	O
C2	O
servers	O
were	O
used	O
in	O
this	O
campaign	O
,	O
but	O
one	O
(	O
37.1.207.31	B-Indicator
)	O
was	O
the	O
most	O
heavily	O
used	O
.	O

Figure	O
7	O
.	O

RuMMS	B-Malware
samples	O
and	O
C2	O
servers	O
Figure	O
8	O
shows	O
how	O
these	O
samples	O
,	O
C2	O
servers	O
and	O
hosting	O
websites	O
are	O
related	O
to	O
each	O
other	O
,	O
including	O
when	O
they	O
were	O
compiled	O
or	O
observed	O
.	O

In	O
the	O
quadrant	O
,	O
the	O
smaller	O
boxes	O
in	O
blue-gray	O
represent	O
particular	O
apps	O
in	O
the	O
RuMMS	B-Malware
family	O
,	O
while	O
the	O
bigger	O
boxes	O
in	O
deep-blue	O
represent	O
C2	O
servers	O
used	O
by	O
some	O
RuMMS	B-Malware
apps	O
.	O

The	O
dotted	O
arrows	O
represent	O
the	O
use	O
of	O
a	O
particular	O
C2	O
server	O
by	O
a	O
specific	O
app	O
to	O
send	O
information	O
and	O
fetch	O
instructions	O
.	O

In	O
this	O
figure	O
we	O
have	O
11	O
RuMMS	B-Malware
samples	O
,	O
all	O
of	O
which	O
were	O
hosted	O
on	O
the	O
website	O
as	O
shown	O
in	O
the	O
“	O
y	O
”	O
axis	O
.	O

The	O
dates	O
on	O
the	O
“	O
x	O
”	O
axis	O
show	O
the	O
dates	O
when	O
we	O
first	O
saw	O
these	O
apps	O
in	O
the	O
wild	O
.	O

This	O
figure	O
demonstrates	O
the	O
following	O
interesting	O
information	O
:	O
The	O
time	O
range	O
when	O
threat	O
actors	O
distributed	O
RuMMS	B-Malware
on	O
those	O
shared-hosting	O
websites	O
is	O
from	O
January	O
2016	O
to	O
March	O
2016	O
.	O

Threat	O
actors	O
used	O
different	O
websites	O
to	O
host	O
different	O
payloads	O
at	O
different	O
times	O
.	O

This	O
kind	O
of	O
“	O
moving	O
target	O
”	O
behavior	O
made	O
it	O
harder	O
to	O
track	O
their	O
actions	O
.	O

The	O
same	O
websites	O
have	O
hosted	O
different	O
RuMMS	B-Malware
samples	O
at	O
different	O
dates	O
.	O

C2	O
servers	O
are	O
shared	O
by	O
multiple	O
samples	O
.	O

This	O
matches	O
our	O
observations	O
of	O
C2	O
servers	O
as	O
shown	O
in	O
Figure	O
7	O
.	O

Figure	O
8	O
.	O

RuMMS	B-Malware
samples	O
,	O
hosting	O
sites	O
,	O
C2	O
servers	O
from	O
Jan.	O
2016	O
to	O
Mar	O
.	O

2016	O
We	O
do	O
not	O
know	O
exactly	O
how	O
many	O
people	O
have	O
been	O
infected	O
with	O
RuMMS	B-Malware
malware	O
;	O
however	O
,	O
our	O
data	O
suggests	O
that	O
there	O
are	O
at	O
least	O
2,729	O
infections	O
with	O
RuMMS	B-Malware
samples	O
from	O
January	O
2016	O
to	O
early	O
April	O
2016	O
.	O

Figure	O
9	O
shows	O
the	O
number	O
of	O
RuMMS	B-Malware
infections	O
recorded	O
in	O
the	O
last	O
four	O
months	O
.	O

When	O
we	O
first	O
observed	O
the	O
malware	O
in	O
January	O
,	O
we	O
recorded	O
380	O
infections	O
.	O

In	O
February	O
,	O
we	O
recorded	O
767	O
infections	O
.	O

In	O
March	O
,	O
it	O
peaked	O
at	O
1,169	O
infections	O
.	O

In	O
April	O
,	O
at	O
the	O
time	O
of	O
writing	O
this	O
post	O
,	O
we	O
recorded	O
413	O
RuMMS	B-Malware
infections	O
.	O

Although	O
the	O
propagation	O
trend	O
seems	O
to	O
be	O
slowing	O
down	O
a	O
bit	O
,	O
the	O
figure	O
tells	O
us	O
that	O
RuMMS	B-Malware
malware	O
is	O
still	O
alive	O
in	O
the	O
wild	O
.	O

We	O
continue	O
to	O
monitor	O
its	O
progress	O
.	O

Conclusion	O
Smishing	O
(	O
SMS	O
phishing	O
)	O
offers	O
a	O
unique	O
vector	O
to	O
infect	O
mobile	O
users	O
.	O

The	O
recent	O
RuMMS	B-Malware
campaign	O
shows	O
that	O
Smishing	O
is	O
still	O
a	O
popular	O
means	O
for	O
threat	O
actors	O
to	O
distribute	O
their	O
malware	O
.	O

In	O
addition	O
,	O
the	O
use	O
of	O
shared-hosting	O
providers	O
adds	O
flexibility	O
to	O
the	O
threat	O
actor	O
’	O
s	O
campaign	O
and	O
makes	O
it	O
harder	O
for	O
defending	O
parties	O
to	O
track	O
these	O
moving	O
targets	O
.	O

Fortunately	O
,	O
FireEye	B-System
Mobile	I-System
Threat	I-System
Prevention	I-System
platform	O
can	O
recognize	O
the	O
malicious	O
SMS	O
and	O
networking	O
behaviors	O
used	O
by	O
these	O
RuMMS	B-Malware
samples	O
,	O
and	O
help	O
us	O
quickly	O
identify	O
the	O
threat	O
.	O

To	O
protect	O
yourself	O
from	O
these	O
threats	O
,	O
FireEye	B-Organization
suggests	O
that	O
users	O
:	O
Take	O
caution	O
before	O
clicking	O
any	O
links	O
where	O
you	O
are	O
not	O
sure	O
about	O
the	O
origin	O
.	O

Don	O
’	O
t	O
install	O
apps	O
outside	O
the	O
official	O
app	O
store	O
.	O

Exodus	B-Malware
:	O
New	O
Android	B-System
Spyware	O
Made	O
in	O
Italy	O
Mar	O
29	O
Summary	O
We	O
identified	O
a	O
new	O
Android	B-System
spyware	O
platform	O
we	O
named	O
Exodus	B-Malware
,	O
which	O
is	O
composed	O
of	O
two	O
stages	O
we	O
call	O
Exodus	B-Malware
One	I-Malware
and	O
Exodus	B-Malware
Two	I-Malware
.	O

We	O
have	O
collected	O
numerous	O
samples	O
spanning	O
from	O
2016	O
to	O
early	O
2019	O
.	O

Instances	O
of	O
this	O
spyware	O
were	O
found	O
on	O
the	O
Google	B-System
Play	I-System
Store	I-System
,	O
disguised	O
as	O
service	O
applications	O
from	O
mobile	O
operators	O
.	O

Both	O
the	O
Google	B-System
Play	I-System
Store	I-System
pages	O
and	O
the	O
decoys	O
of	O
the	O
malicious	O
apps	O
are	O
in	O
Italian	O
.	O

According	O
to	O
publicly	O
available	O
statistics	O
,	O
as	O
well	O
as	O
confirmation	O
from	O
Google	B-Organization
,	O
most	O
of	O
these	O
apps	O
collected	O
a	O
few	O
dozens	O
installations	O
each	O
,	O
with	O
one	O
case	O
reaching	O
over	O
350	O
.	O

All	O
of	O
the	O
victims	O
are	O
located	O
in	O
Italy	O
.	O

All	O
of	O
these	O
Google	B-System
Play	I-System
Store	I-System
pages	O
have	O
been	O
taken	O
down	O
by	O
Google	B-Organization
.	O

We	O
believe	O
this	O
spyware	O
platform	O
is	O
developed	O
by	O
an	O
Italian	O
company	O
called	O
eSurv	B-Organization
,	O
which	O
primarily	O
operates	O
in	O
the	O
business	O
of	O
video	O
surveillance	O
.	O

According	O
to	O
public	O
records	O
it	O
appears	O
that	O
eSurv	B-Organization
began	O
to	O
also	O
develop	O
intrusion	O
software	O
in	O
2016	O
.	O

Exodus	B-Malware
is	O
equipped	O
with	O
extensive	O
collection	O
and	O
interception	O
capabilities	O
.	O

Worryingly	O
,	O
some	O
of	O
the	O
modifications	O
enforced	O
by	O
the	O
spyware	O
might	O
expose	O
the	O
infected	O
devices	O
to	O
further	O
compromise	O
or	O
data	O
tampering	O
.	O

Disguised	O
Spyware	O
Uploaded	O
on	O
Google	B-System
Play	I-System
Store	I-System
We	O
identified	O
previously	O
unknown	O
spyware	O
apps	O
being	O
successfully	O
uploaded	O
on	O
Google	B-System
Play	I-System
Store	I-System
multiple	O
times	O
over	O
the	O
course	O
of	O
over	O
two	O
years	O
.	O

These	O
apps	O
would	O
remain	O
available	O
on	O
the	O
Play	B-System
Store	I-System
for	O
months	O
and	O
would	O
eventually	O
be	O
re-uploaded	O
.	O

While	O
details	O
would	O
vary	O
,	O
all	O
of	O
the	O
identified	O
copies	O
of	O
this	O
spyware	O
shared	O
a	O
similar	O
disguise	O
.	O

In	O
most	O
cases	O
they	O
would	O
be	O
crafted	O
to	O
appear	O
as	O
applications	O
distributed	O
by	O
unspecified	O
mobile	O
operators	O
in	O
Italy	O
.	O

Often	O
the	O
app	O
description	O
on	O
the	O
Play	B-System
Store	I-System
would	O
reference	O
some	O
SMS	O
messages	O
the	O
targets	O
would	O
supposedly	O
receive	O
leading	O
them	O
to	O
the	O
Play	B-System
Store	I-System
page	O
.	O

All	O
of	O
the	O
Play	B-System
Store	I-System
pages	O
we	O
identified	O
and	O
all	O
of	O
the	O
decoys	O
of	O
the	O
apps	O
themselves	O
are	O
written	O
in	O
Italian	O
.	O

According	O
to	O
Google	B-Organization
,	O
whom	O
we	O
have	O
contacted	O
to	O
alert	O
about	O
our	O
discoveries	O
,	O
nearly	O
25	O
variants	O
of	O
this	O
spyware	O
were	O
uploaded	O
on	O
Google	B-System
Play	I-System
Store	I-System
.	O

Google	B-System
Play	I-System
has	O
removed	O
the	O
apps	O
and	O
they	O
stated	O
that	O
"	O
thanks	O
to	O
enhanced	O
detection	O
models	O
,	O
Google	B-System
Play	I-System
Protect	I-System
will	O
now	O
be	O
able	O
to	O
better	O
detect	O
future	O
variants	O
of	O
these	O
applications	O
''	O
.	O

While	O
Google	I-Organization
did	O
not	O
share	O
with	O
us	O
the	O
total	O
number	O
of	O
infected	O
devices	O
,	O
they	O
confirmed	O
that	O
one	O
of	O
these	O
malicious	O
apps	O
collected	O
over	O
350	O
installations	O
through	O
the	O
Play	B-System
Store	I-System
,	O
while	O
other	O
variants	O
collected	O
few	O
dozens	O
each	O
,	O
and	O
that	O
all	O
infections	O
were	O
located	O
in	O
Italy	O
.	O

We	O
have	O
directly	O
observed	O
multiple	O
copies	O
of	O
Exodus	B-Malware
with	O
more	O
than	O
50	O
installs	O
and	O
we	O
can	O
estimate	O
the	O
total	O
number	O
of	O
infections	O
to	O
amount	O
in	O
the	O
several	O
hundreds	O
,	O
if	O
not	O
a	O
thousand	O
or	O
more	O
.	O

Stage	O
1	O
:	O
Exodus	B-Malware
One	I-Malware
The	O
first	O
stage	O
installed	O
by	O
downloading	O
the	O
malicious	O
apps	O
uploaded	O
on	O
Google	B-System
Play	I-System
Store	I-System
only	O
acts	O
as	O
a	O
dropper	O
.	O

Following	O
are	O
some	O
examples	O
of	O
the	O
decoys	O
used	O
by	O
these	O
droppers	O
:	O
The	O
purpose	O
of	O
Exodus	B-Malware
One	I-Malware
seems	O
to	O
be	O
to	O
collect	O
some	O
basic	O
identifying	O
information	O
about	O
the	O
device	O
(	O
namely	O
the	O
IMEI	O
code	O
and	O
the	O
phone	O
number	O
)	O
and	O
send	O
it	O
to	O
the	O
Command	O
&	O
Control	O
server	O
.	O

This	O
is	O
usually	O
done	O
in	O
order	O
to	O
validate	O
the	O
target	O
of	O
a	O
new	O
infection	O
.	O

This	O
is	O
further	O
corroborated	O
by	O
some	O
older	O
and	O
unobfuscated	O
samples	O
from	O
2016	O
,	O
whose	O
primary	O
classes	O
are	O
named	O
CheckValidTarget	O
.	O

During	O
our	O
tests	O
the	O
spyware	O
was	O
upgraded	O
to	O
the	O
second	O
stage	O
on	O
our	O
test	O
device	O
immediately	O
after	O
the	O
first	O
check-ins	O
.	O

This	O
suggests	O
that	O
the	O
operators	O
of	O
the	O
Command	O
&	O
Control	O
are	O
not	O
enforcing	O
a	O
validation	O
of	O
the	O
targets	O
.	O

Additionally	O
,	O
during	O
a	O
period	O
of	O
several	O
days	O
,	O
our	O
infected	O
test	O
device	O
was	O
never	O
remotely	O
disinfected	O
by	O
the	O
operators	O
.	O

For	O
the	O
purpose	O
of	O
this	O
report	O
we	O
analyze	O
here	O
the	O
Exodus	B-Malware
One	I-Malware
sample	O
with	O
hash	O
8453ce501fee1ca8a321f16b09969c517f92a24b058ac5b54549eabd58bf1884	B-Indicator
which	O
communicated	O
with	O
the	O
Command	O
&	O
Control	O
server	O
at	O
54.71.249.137	B-Indicator
.	O

Other	O
samples	O
communicated	O
with	O
other	O
servers	O
listed	O
at	O
the	O
bottom	O
of	O
this	O
report	O
.	O

Exodus	O
One	O
checks-in	O
by	O
sending	O
a	O
POST	O
request	O
containing	O
the	O
app	O
package	O
name	O
,	O
the	O
device	O
IMEI	O
and	O
an	O
encrypted	O
body	O
containing	O
additional	O
device	O
information	O
.	O

The	O
encrypted	O
body	O
is	O
composed	O
of	O
various	O
identifiers	O
which	O
are	O
joined	O
together	O
:	O
doFinal	O
(	O
)	O
is	O
called	O
to	O
encrypt	O
the	O
device	O
information	O
string	O
:	O
The	O
user	O
agent	O
string	O
is	O
built	O
from	O
the	O
package	O
name	O
and	O
IMEI	O
number	O
:	O
Finally	O
the	O
HTTP	O
request	O
is	O
sent	O
to	O
the	O
server	O
at	O
https	B-Indicator
:	I-Indicator
//54.71.249.137/eddd0317-2bdc-4140-86cb-0e8d7047b874	I-Indicator
.	O

Many	O
of	O
the	O
strings	O
in	O
the	O
application	O
are	O
XOR	O
'd	O
with	O
the	O
key	O
Kjk1MmphFG	O
:	O
After	O
some	O
additional	O
requests	O
,	O
the	O
dropper	O
made	O
a	O
POST	O
request	O
to	O
https	B-Indicator
:	I-Indicator
//54.71.249.137/56e087c9-fc56-49bb-bbd0-4fafc4acd6e1	I-Indicator
which	O
returned	O
a	O
zip	O
file	O
containing	O
the	O
second	O
stage	O
binaries	O
.	O

Stage	O
2	O
:	O
Exodus	B-Malware
Two	I-Malware
The	O
Zip	O
archive	O
returned	O
by	O
the	O
check-in	O
performed	O
by	O
Exodus	B-Malware
One	I-Malware
is	O
a	O
collection	O
of	O
files	O
including	O
the	O
primary	O
payload	O
mike.jar	B-Indicator
and	O
several	O
compiled	O
utilities	O
that	O
serve	O
different	O
functions	O
.	O

At	O
least	O
in	O
most	O
recent	O
versions	O
,	O
as	O
of	O
January	O
2019	O
,	O
the	O
Zip	O
archive	O
would	O
actually	O
contain	O
the	O
i686	O
,	O
arm	O
and	O
arm64	O
versions	O
of	O
all	O
deployed	O
binaries	O
.	O

File	O
Name	O
Modified	O
Date	O
SHA256	O
null_arm	O
2018-02-27	O
06:44:00	O
48a7dd672931e408662d2b5e1abcd6ef00097b8ffe3814f0d2799dd6fd74bd88	B-Indicator
null_i686	O
2018-02-27	O
06:44:00	O
c228a534535b22a316a97908595a2d793d0fecabadc32846c6d1bfb08ca9a658	B-Indicator
null_arm64	O
2018-02-27	O
06:43:00	O
48a7dd672931e408662d2b5e1abcd6ef00097b8ffe3814f0d2799dd6fd74bd88	B-Indicator

sepolicy-inject_arm	O
2019-01-08	O
04:55:00	O
47449a612697ad99a6fbd6e02a84e957557371151f2b034a411ebb10496648c8	B-Indicator
sepolicy-inject_arm64	O
2019-01-08	O
04:55:00	O
824ad333320cbb7873dc49e61c14f749b0e0d88723635524463f2e6f56ea133a	B-Indicator
sepolicy-inject_i686	O
2019-01-08	O
04:55:00	O
13ec6cec511297ac3137cf7d6e4a7c4f5dd2b24478a06262a44f13a3d61070b6	B-Indicator

rootdaemon_arm	O
2019-01-08	O
04:55:00	O
00c787c0c0bc26caf623e66373a5aaa1b913b9caee1f34580bdfdd21954b7cc4	B-Indicator
rootdaemon_arm64	O
2019-01-08	O
04:55:00	O
3ee3a973c62ba5bd9eab595a7c94b7a26827c5fa5b21964d511ab58903929ec5	B-Indicator
mike.jar	B-Indicator
2018-12-06	O
05:50:00	O
a42a05bf9b412cd84ea92b166d790e8e72f1d01764f93b05ace62237fbabe40e	B-Indicator

rootdaemon_i686	O
2019-01-08	O
04:55:00	O
b46f282f9a1bce3798faee3212e28924730a657eb93cda3824c449868b6ee2e7	B-Indicator
zygotedaemonarm	O
2019-01-08	O
04:55:00	O
e3f65f84dd6c2c3a5a653a3788d78920c0321526062a6b53daaf23fa57778a5f	B-Indicator
zygotedaemonarm64	O
2019-01-08	O
04:55:00	O
11499ff2418f4523344de81a447f6786fdba4982057d4114f64db929990b4b59	B-Indicator

zygotedaemoni686	O
2019-01-08	O
04:55:00	O
3c9f08b3280851f54414dfa5a57f40d3b7be7b73736fa0ba21b078e75ce54d33	B-Indicator
sapp.apk	B-Indicator
2019-01-08	O
04:53:00	O
4bf1446c412dd5c552539490d03e999a6ceb96ae60a9e7846427612bec316619	B-Indicator
placeholder	O
2018-03-29	O
16:31:00	O
e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855	B-Indicator

After	O
download	O
,	O
Exodus	B-Malware
One	I-Malware
would	O
dynamically	O
load	O
and	O
execute	O
the	O
primary	O
stage	O
2	O
payload	O
mike.jar	B-Indicator
using	O
the	O
Android	B-System
API	I-System
DexClassLoader	O
(	O
)	O
.	O

mike.jar	B-Indicator
implements	O
most	O
of	O
the	O
data	O
collection	O
and	O
exfiltration	O
capabilities	O
of	O
this	O
spyware	O
.	O

Of	O
the	O
various	O
binaries	O
downloaded	O
,	O
the	O
most	O
interesting	O
are	O
null	O
,	O
which	O
serves	O
as	O
a	O
local	O
and	O
reverse	O
shell	O
,	O
and	O
rootdaemon	O
,	O
which	O
takes	O
care	O
of	O
privilege	O
escalation	O
and	O
data	O
acquisition	O
.	O

rootdaemon	O
will	O
first	O
attempt	O
to	O
jailbreak	O
the	O
device	O
using	O
a	O
modified	O
version	O
of	O
the	O
DirtyCow	B-Vulnerability
exploit	I-Vulnerability
.	O

Similarly	O
to	O
another	O
Android	B-System
spyware	O
made	O
in	O
Italy	O
,	O
originally	O
discovered	O
by	O
Lukas	O
Stefanko	O
and	O
later	O
named	O
Skygofree	B-Malware
and	O
analyzed	O
in	O
depth	O
by	O
Kaspersky	B-Organization
Labs	I-Organization
,	O
Exodus	B-Malware
also	O
takes	O
advantage	O
of	O
"	O
protectedapps	O
''	O
,	O
a	O
feature	O
in	O
Huawei	B-Organization
phones	O
that	O
allows	O
to	O
configure	O
power-saving	O
options	O
for	O
running	O
applications	O
.	O

By	O
manipulating	O
a	O
SQLite	O
database	O
,	O
Exodus	B-Malware
is	O
able	O
to	O
keep	O
itself	O
running	O
even	O
when	O
the	O
screen	O
goes	O
off	O
and	O
the	O
application	O
would	O
otherwise	O
be	O
suspended	O
to	O
reduce	O
battery	O
consumption	O
.	O

Additionally	O
,	O
rootdaemon	O
attempts	O
to	O
remove	O
its	O
own	O
power	O
usage	O
statistics	O
from	O
Huawei	B-Organization
phones	O
'	O
SystemManager	O
:	O
Similarly	O
,	O
the	O
malicious	O
application	O
probably	O
attempts	O
to	O
minimize	O
traces	O
on	O
Samsung	B-Organization
phones	O
by	O
adding	O
to	O
the	O
file	O
/data/data/com.samsung.android.securitylogagent/shared_prefs/apm_sp_status_of_apps.xml	B-Indicator
the	O
following	O
lines	O
:	O
And	O
adding	O
to	O
the	O
file	O
/data/data/com.samsung.android.securitylogagent/shared_prefs/com.samsung.android.securitylogagent_preferences.xml	B-Indicator

these	O
lines	O
instead	O
:	O
Data	O
Collection	O
and	O
Exfiltration	O
As	O
mentioned	O
,	O
mike.jar	B-Indicator
equips	O
the	O
spyware	O
with	O
extensive	O
collection	O
capabilities	O
,	O
including	O
:	O
Retrieve	O
a	O
list	O
of	O
installed	O
applications	O
.	O

Record	O
surroundings	O
using	O
the	O
built-in	O
microphone	O
in	O
3gp	O
format	O
.	O

Retrieve	O
the	O
browsing	O
history	O
and	O
bookmarks	O
from	O
Chrome	B-System
and	O
SBrowser	B-System
(	O
the	O
browser	O
shipped	O
with	O
Samsung	B-Organization
phones	O
)	O
.	O

Extract	O
events	O
from	O
the	O
Calendar	B-System
app	I-System
.	O

Extract	O
the	O
calls	O
log	O
.	O

Record	O
phone	O
calls	O
audio	O
in	O
3gp	O
format	O
.	O

Take	O
pictures	O
with	O
the	O
embedded	O
camera	O
.	O

Collect	O
information	O
on	O
surrounding	O
cellular	O
towers	O
(	O
BTS	O
)	O
.	O

Extract	O
the	O
address	B-System
book	I-System
.	O

Extract	O
the	O
contacts	O
list	O
from	O
the	O
Facebook	B-System
app	I-System
.	O

Extract	O
logs	O
from	O
Facebook	B-System
Messenger	I-System
conversations	O
.	O

Take	O
a	O
screenshot	O
of	O
any	O
app	O
in	O
foreground	O
.	O

Extract	O
information	O
on	O
pictures	O
from	O
the	O
Gallery	O
.	O

Extract	O
information	O
from	O
th	O
GMail	B-System
app	O
.	O

Dump	O
data	O
from	O
the	O
IMO	O
messenger	B-System
app	O
.	O

Extract	O
call	O
logs	O
,	O
contacts	O
and	O
messages	O
from	O
the	O
Skype	B-System
app	O
.	O

Retrieve	O
all	O
SMS	O
messages	O
.	O

Extract	O
messages	O
and	O
the	O
encryption	O
key	O
from	O
the	O
Telegram	B-System
app	O
.	O

Dump	O
data	O
from	O
the	O
Viber	B-System
messenger	I-System
app	O
.	O

Extract	O
logs	O
from	O
WhatsApp	B-System
.	O

Retrieve	O
media	O
exchanged	O
through	O
WhatsApp	B-System
.	O

Extract	O
the	O
Wi-Fi	O
network	O
's	O
password	O
.	O

Extract	O
data	O
from	O
WeChat	B-System
app	O
.	O

Extract	O
current	O
GPS	O
coordinates	O
of	O
the	O
phone	O
.	O

While	O
some	O
of	O
these	O
acquisition	O
are	O
performed	O
purely	O
through	O
code	O
in	O
mike.jar	B-Indicator
,	O
some	O
others	O
that	O
require	O
access	O
to	O
,	O
for	O
example	O
,	O
SQLite	O
databases	O
or	O
other	O
files	O
in	O
the	O
application	O
's	O
storage	O
are	O
performed	O
through	O
rootdaemon	O
instead	O
,	O
which	O
should	O
be	O
running	O
with	O
root	O
privileges	O
.	O

In	O
order	O
to	O
achieve	O
this	O
,	O
mike.jar	B-Indicator
connects	O
to	O
rootdaemon	O
through	O
various	O
TCP	O
ports	O
that	O
the	O
daemon	O
binds	O
on	O
some	O
extraction	O
routines	O
for	O
supported	O
applications	O
:	O
Port	B-Indicator
6202	I-Indicator
:	O
WhatsApp	B-System
extraction	O
service	O
.	O

Ports	B-Indicator
6203	I-Indicator
and	I-Indicator
6204	I-Indicator
:	O
Facebook	B-Organization
extraction	O
service	O
.	O

Port	B-Indicator
6205	I-Indicator
:	O
Gmail	B-System
extraction	O
service	O
.	O

Port	B-Indicator
6206	I-Indicator
:	O
Skype	B-System
extraction	O
service	O
.	O

Port	B-Indicator
6207	I-Indicator
:	O
Viber	B-System
extraction	O
service	O
.	O

Port	B-Indicator
6208	I-Indicator
:	O
IMO	B-System
extraction	O
service	O
.	O

Port	B-Indicator
6209	I-Indicator
:	O
Telegram	B-System
extraction	O
service	O
.	O

Port	B-Indicator
6210	I-Indicator
:	O
SBrowser	B-System
extraction	O
service	O
.	O

Port	B-Indicator
6211	I-Indicator
:	O
Calendar	B-System
extraction	O
service	O
.	O

Port	B-Indicator
6212	I-Indicator
:	O
Chrome	B-System
extraction	O
service	O
.	O

These	O
services	O
appear	O
to	O
be	O
running	O
on	O
all	O
network	O
interfaces	O
and	O
are	O
therefore	O
accessible	O
to	O
anyone	O
sharing	O
a	O
local	O
network	O
with	O
an	O
infected	O
device	O
.	O

Following	O
we	O
can	O
see	O
an	O
example	O
of	O
a	O
connection	O
to	O
port	B-Indicator
6209	I-Indicator
which	O
is	O
used	O
to	O
extract	O
data	O
from	O
the	O
Telegram	B-System
app	O
.	O

We	O
are	O
able	O
to	O
send	O
commands	O
to	O
the	O
service	O
such	O
as	O
dumpmsgdb	O
or	O
getkey	O
(	O
which	O
dumps	O
the	O
tgnet.dat	B-Indicator
file	I-Indicator
)	O
.	O

Data	O
acquired	O
from	O
mike.jar	B-Indicator
's	O
extraction	O
modules	O
is	O
normally	O
XORed	O
and	O
stored	O
in	O
a	O
folder	O
named	O
.lost+found	O
on	O
the	O
SD	O
card	O
.	O

Data	O
is	O
eventually	O
exfiltrated	O
over	O
a	O
TLS	O
connection	O
to	O
the	O
Command	O
&	O
Control	O
server	B-Indicator
ws.my-local-weather	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
through	O
an	O
upload	O
queue	O
.	O

As	O
mentioned	O
before	O
,	O
our	O
test	O
device	O
was	O
automatically	O
from	O
stage	O
one	O
to	O
stage	O
two	O
,	O
which	O
started	O
collecting	O
data	O
.	O

For	O
example	O
,	O
the	O
password	O
of	O
the	O
WiFi	O
network	O
used	O
by	O
the	O
phone	O
was	O
stored	O
in	O
the	O
folder	O
/storage/emulated/0/.lost+found/0BBDA068-9D27-4B55-B226-299FCF2B4242/	B-Indicator
using	O
the	O
following	O
file	O
name	O
format	O
DD_MM_2019_HH_mm_ss_XXXXXXXXXXXXX.txt.crypt	B-Indicator
(	O
the	O
datetime	O
followed	O
by	O
the	O
IMEI	O
)	O
.	O

Eventually	O
we	O
observed	O
the	O
agent	O
exfiltrate	O
the	O
WiFi	O
password	O
from	O
our	O
test	O
phone	O
to	O
the	O
Command	O
&	O
Control	O
server	O
:	O
Similarly	O
,	O
the	O
agent	O
also	O
sent	O
to	O
the	O
Command	O
&	O
Control	O
the	O
list	O
of	O
installed	O
apps	O
:	O
This	O
Command	O
&	O
Control	O
seems	O
to	O
have	O
been	O
active	O
since	O
at	O
least	O
April	O
2017	O
and	O
was	O
registered	O
impersonating	O
the	O
legitimate	O
service	O
AccuWeather	B-System
.	O

Local	O
and	O
Remote	O
Shells	O
In	O
order	O
to	O
execute	O
commands	O
on	O
the	O
infected	O
devices	O
,	O
as	O
well	O
as	O
to	O
provide	O
a	O
reverse	O
shell	O
to	O
the	O
Command	O
&	O
Control	O
operators	O
,	O
Exodus	B-Malware
Two	I-Malware
immediately	O
attempts	O
to	O
execute	O
a	O
payload	O
it	O
downloads	O
with	O
the	O
name	O
null	O
.	O

Once	O
launched	O
,	O
null	O
will	O
first	O
verify	O
whether	O
it	O
is	O
able	O
to	O
fork	O
on	O
the	O
system	O
and	O
that	O
there	O
is	O
no	O
other	O
instance	O
of	O
itself	O
currently	O
running	O
by	O
checking	O
whether	O
the	O
local	O
port	B-Indicator
number	I-Indicator
6842	I-Indicator
is	O
available	O
.	O

This	O
payload	O
will	O
then	O
attempt	O
to	O
instantiate	O
a	O
remote	O
reverse	O
/system/bin/sh	B-Indicator
shell	O
to	O
the	O
Command	O
&	O
Control	O
ws.my-local-weather	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
on	O
port	B-Indicator
22011	I-Indicator
.	O

It	O
is	O
worth	O
noticing	O
that	O
this	O
remote	O
reverse	O
shell	O
does	O
not	O
employ	O
any	O
transport	O
cryptography	O
.	O

The	O
traffic	O
transits	O
in	O
clear	O
and	O
is	O
therefore	O
potentially	O
exposed	O
to	O
man-in-the-middle	O
attacks	O
:	O
At	O
the	O
same	O
time	O
,	O
null	O
will	O
also	O
bind	O
a	O
local	O
shell	O
on	O
0.0.0.0:6842	B-Indicator
.	O

This	O
local	O
port	O
is	O
used	O
by	O
Exodus	B-Malware
Two	I-Malware
to	O
execute	O
various	O
commands	O
on	O
the	O
Android	B-System
device	O
,	O
such	O
as	O
enabling	O
or	O
disabling	O
certain	O
services	O
,	O
or	O
parsing	O
app	O
databases	O
.	O

However	O
,	O
binding	O
a	O
shell	O
on	O
all	O
available	O
interfaces	O
will	O
obviously	O
make	O
it	O
accessible	O
to	O
anyone	O
who	O
is	O
sharing	O
at	O
least	O
a	O
local	O
network	O
with	O
an	O
infected	O
device	O
.	O

For	O
example	O
,	O
if	O
an	O
infected	O
device	O
is	O
connected	O
to	O
a	O
public	O
Wi-Fi	O
network	O
any	O
other	O
host	O
will	O
be	O
able	O
to	O
obtain	O
a	O
terminal	O
on	O
the	O
device	O
without	O
any	O
form	O
of	O
authentication	O
or	O
verification	O
by	O
simply	O
connecting	O
to	O
the	O
port	O
.	O

If	O
the	O
mobile	O
operator	O
does	O
n't	O
enforce	O
proper	O
client	O
isolation	O
,	O
it	O
is	O
possible	O
that	O
the	O
infected	O
devices	O
are	O
also	O
exposed	O
to	O
the	O
rest	O
of	O
the	O
cellular	O
network	O
.	O

Obviously	O
,	O
this	O
inevitably	O
leaves	O
the	O
device	O
open	O
not	O
only	O
to	O
further	O
compromise	O
but	O
to	O
data	O
tampering	O
as	O
well	O
.	O

null	O
is	O
not	O
the	O
only	O
payload	O
opening	O
a	O
shell	O
on	O
the	O
phone	O
.	O

The	O
rootdaemon	O
binary	O
in	O
fact	O
offers	O
several	O
other	O
possibilities	O
to	O
execute	O
commands	O
on	O
the	O
infected	O
device	O
just	O
by	O
connecting	O
to	O
TCP	O
port	B-Indicator
6200	I-Indicator
and	O
issuing	O
one	O
of	O
the	O
following	O
commands	O
.	O

Sending	O
the	O
command	O
sh	O
to	O
TCP	O
port	B-Indicator
6200	I-Indicator
results	O
in	O
a	O
full	O
terminal	O
being	O
dropped	O
:	O
Sending	O
the	O
command	O
cmd	O
followed	O
by	O
a	O
proper	O
terminal	O
command	O
will	O
execute	O
it	O
and	O
print	O
the	O
output	O
(	O
in	O
the	O
example	O
we	O
use	O
id	O
which	O
displays	O
the	O
identity	O
of	O
the	O
system	O
user	O
running	O
the	O
issued	O
commands	O
)	O
:	O
Doing	O
the	O
same	O
as	O
above	O
but	O
with	O
command	O
sucmd	O
will	O
run	O
the	O
terminal	O
command	O
as	O
root	O
:	O
Other	O
commands	O
supported	O
by	O
rootdaemon	O
on	O
TCP	O
port	B-Indicator
6200	I-Indicator
are	O
su	O
(	O
which	O
in	O
our	O
tests	O
did	O
n't	O
properly	O
work	O
)	O
,	O
loadsocketpolicy	O
,	O
loadfilepolicy	O
,	O
remount	O
and	O
removeroot	O

.	O

At	O
the	O
cost	O
of	O
possibly	O
being	O
overly	O
verbose	O
,	O
following	O
is	O
the	O
output	O
of	O
an	O
nmap	O
scan	O
of	O
the	O
infected	O
Android	O
device	O
from	O
a	O
laptop	O
in	O
the	O
same	O
local	O
network	O
,	O
which	O
further	O
demonstrantes	O
the	O
availability	O
of	O
the	O
same	O
open	O
TCP	O
ports	O
that	O
we	O
have	O
mentioned	O
thus	O
far	O
:	O
Identification	O
of	O
eSurv	B-Organization
Presence	O
of	O
Italian	O
language	O
At	O
a	O
first	O
look	O
,	O
the	O
first	O
samples	O
of	O
the	O
spyware	O
we	O
obtained	O
did	O
not	O
show	O
immediately	O
evident	O
connections	O
to	O
any	O
company	O
.	O

However	O
,	O
the	O
persistent	O
presence	O
of	O
Italian	O
language	O
both	O
on	O
the	O
Google	B-System
Play	I-System
Store	O
pages	O
as	O
well	O
as	O
inside	O
the	O
spyware	O
code	O
was	O
a	O
clear	O
sign	O
that	O
an	O
Italian	O
actor	O
was	O
behind	O
the	O
creation	O
of	O
this	O
platform	O
.	O

Initially	O
some	O
particular	O
words	O
from	O
the	O
decompiled	O
classes.dex	B-Indicator
of	O
Exodus	B-Malware
Two	O
sent	O
us	O
in	O
the	O
right	O
direction	O
.	O

"	O
Mundizza	O
''	O
is	O
a	O
dialectal	O
word	O
,	O
a	O
derivative	O
of	O
the	O
proper	O
Italian	O
word	O
"	O
immondizia	O
''	O
that	O
translates	O
to	O
"	O
trash	O
''	O
or	O
"	O
garbage	O
''	O
in	O
English	O
.	O

Interestingly	O
,	O
"	O
mundizza	O
''	O
is	O
typical	O
of	O
Calabria	O
,	O
a	O
region	O
in	O
the	O
south	O
of	O
Italy	O
,	O
and	O
more	O
specifically	O
it	O
appears	O
to	O
be	O
language	O
native	O
of	O
the	O
city	O
of	O
Catanzaro	O
.	O

Additionally	O
,	O
some	O
copies	O
of	O
Exodus	O
One	O
use	O
the	O
following	O
XOR	O
key	O
:	O
Rino	O
Gattuso	O
is	O
a	O
famous	O
retired	O
Italian	O
footballer	O
,	O
originally	O
from	O
Calabria	O
.	O

While	O
not	O
too	O
seriously	O
,	O
these	O
elements	O
made	O
us	O
restrict	O
our	O
research	O
into	O
surveillance	O
companies	O
from	O
the	O
region	O
.	O

Overlapping	O
Infrastructure	O
with	O
eSurv	O
Surveillance	O
Cameras	O
The	O
Command	O
&	O
Control	O
domain	O
configured	O
in	O
several	O
of	O
the	O
malicious	O
applications	O
found	O
on	O
Google	B-System
Play	I-System
Store	I-System
,	O
ws.my-local-weather	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
,	O
points	O
to	O
the	O
IP	O
address	O
54.69.156.31	B-Indicator
which	O
serves	O
a	O
self-signed	O
TLS	O
certificate	O
with	O
the	O
certificate	O
common	O
name	O
MyCert	O
and	O
fingerprint	O
11:41:45:2F	B-Indicator
:	I-Indicator
A7:07:23:54	I-Indicator
:	O
AE:9A	B-Indicator
:	I-Indicator
CE	I-Indicator
:	I-Indicator
F4	I-Indicator
:	I-Indicator
FE:56	I-Indicator
:	I-Indicator
AE	I-Indicator
:	I-Indicator
AC	I-Indicator
:	O
B1	B-Indicator
:	I-Indicator
C2:15:9F:6A	I-Indicator
:	I-Indicator
FC:1E	I-Indicator
:	I-Indicator
CC:7D	I-Indicator
:	B-Indicator
F8:61	I-Indicator
:	I-Indicator
E3:25:26:73:6A	I-Indicator
.	O

A	O
search	O
for	O
this	O
certificate	O
fingerprint	O
on	O
the	O
Internet	O
scanning	O
service	O
Censys	O
returns	O
8	O
additional	O
servers	O
:	O
IP	O
address	O
34.208.71.9	B-Indicator
34.212.92.0	B-Indicator
34.216.43.114	B-Indicator
52.34.144.229	B-Indicator
54.69.156.31	B-Indicator
54.71.249.137	B-Indicator
54.189.5.198	B-Indicator
78.5.0.195	B-Indicator
207.180.245.74	B-Indicator
Opening	O
the	O
Command	O
&	O
Control	O
web	O
page	O
in	O
a	O
browser	O
presents	O
a	O
Basic	O
Authentication	O
prompt	O
:	O
Closing	O
this	O
prompt	O
causes	O
the	O
server	O
to	O
send	O
a	O
"	O
401	O
Unauthorized	O
Response	O
''	O
with	O
an	O
"	O
Access	O
Denied	O
''	O
message	O
in	O
Italian	O

.	O

All	O
of	O
the	O
other	O
IP	O
address	O
we	O
discovered	O
sharing	O
the	O
same	O
TLS	O
certificate	O
behave	O
in	O
the	O
same	O
way	O
.	O

The	O
Command	O
&	O
Control	O
server	O
also	O
displays	O
a	O
favicon	O
image	O
which	O
looks	O
like	O
a	O
small	O
orange	O
ball	O
.	O

At	O
the	O
time	O
of	O
writing	O
,	O
a	O
reverse	O
image	O
search	O
for	O
the	O
favicon	O
on	O
Shodan	O
using	O
the	O
query	O
http.favicon.hash:990643579	B-Indicator
returned	O
around	O
40	O
web	O
servers	O
which	O
use	O
the	O
same	O
favicon	O
.	O

Many	O
of	O
these	O
servers	O
are	O
control	O
panels	O
for	O
video	O
surveillance	O
systems	O
developed	O
by	O
the	O
Italian	O
company	O
eSurv	O
,	O
based	O
in	O
Catanzaro	O
,	O
in	O
Calabria	O
,	O
Italy	O
.	O

Their	O
publicly	O
advertised	O
products	O
include	O
CCTV	O
management	O
systems	O
,	O
surveillance	O
drones	O
,	O
face	O
and	O
license	O
plate	O
recognition	O
systems	O
.	O

eSurv	B-Organization
's	O
logo	O
is	O
identical	O
to	O
the	O
Command	O
&	O
Control	O
server	O
favicon	O
.	O

Older	O
samples	O
connecting	O
to	O
eSurv	B-Organization
Finally	O
,	O
Google	B-Organization
shared	O
with	O
us	O
some	O
older	O
samples	O
of	O
Exodus	B-Malware
One	I-Malware
(	O
with	O
hashes	O
2055584625d24687bd027a63bc0b8faa7d1a854a535de74afba24840a52b1d2f	B-Indicator
and	O
a37f5d2418c5f2f64d06ba28fe62edee1293a56158ddfa9f04020e316054363f	B-Indicator
)	O
which	O
are	O
not	O
obfuscated	O
and	O
use	O
the	O
following	O
disguise	O
:	O
The	O
configuration	O
of	O
these	O
older	O
samples	O

is	O
very	O
similar	O
to	O
newer	O
ones	O
,	O
but	O
it	O
provides	O
additional	O
insights	O
being	O
not	O
obfuscated	O
:	O
Firstly	O
we	O
can	O
notice	O
that	O
,	O
instead	O
of	O
generic	O
domain	O
names	O
or	O
IP	O
addresses	O
,	O
these	O
samples	O
communicated	O
with	O
a	O
Command	O
&	O
Control	O
server	O
located	O
at	O
attiva.exodus.esurv	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
(	O
"	O
attiva	O
''	O
is	O
the	O
Italian	O
for	O
"	O
activate	O
''	O
)	O
.	O

(	O
We	I-System
named	O
the	O
spyware	O
"	O
Exodus	O
''	O
after	O
this	O
Command	O
&	O
Control	O
domain	O
name	O
.	O

)	O
Following	O
is	O
the	O
snippet	O
of	O
code	O
in	O
these	O
older	O
Exodus	B-Malware
One	I-Malware
samples	O
showing	O
the	O
connection	O
to	O
the	O
Command	O
&	O
Control	O
:	O
Below	O
is	O
the	O
almost	O
identical	O
composition	O
of	O
the	O
request	O
to	O
the	O
Command	O
&	O
Control	O
server	O
in	O
mike.jar	B-Indicator
(	O
also	O
containing	O
the	O
path	O
7e661733-e332-429a-a7e2-23649f27690f	O
)	O
:	O
To	O
further	O
corroborate	O
the	O
connection	O
of	O
the	O
Exodus	B-Malware
spyware	I-Malware
with	O
eSurv	O
,	O
the	O
domain	B-Indicator
attiva.exodus.esurv.it	I-Indicator
resolves	O
to	O
the	O
IP	O
212.47.242.236	B-Indicator
which	O
,	O
according	O
to	O

public	O
passive	O
DNS	O
data	O
,	O
in	O
2017	O
was	O
used	O
to	O
host	O
the	O
domain	B-Indicator
server1cs.exodus.connexxa.it	I-Indicator
.	O

Connexxa	O
was	O
a	O
company	O
also	O
from	O
Catanzaro	O
.	O

According	O
to	O
publicly	O
available	O
information	O
,	O
the	O
founder	O
of	O
Connexxa	B-Organization
seems	O
to	O
also	O
be	O
the	O
CEO	O
of	O
eSurv	B-Organization
.	O

Interestingly	O
,	O
we	O
found	O
other	O
DNS	O
records	O
mostly	O
from	O
2017	O
that	O
follow	O
a	O
similar	O
pattern	O
and	O
appear	O
to	O
contain	O
two-letters	O
codes	O
for	O
districts	O
in	O
Italy	O
:	O
Server	O
City	O
server1bo.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Bologna	O
server1bs.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Brescia	O
server1cs.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Cosenza	O
server1ct.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Catania	O
server1fermo.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
server1fi.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Firenze	O
server1gioiat.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
server1na.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Napoli	O
server1rc.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Reggio	O
Calabria	O
server2ct.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Catania	O
server2cz.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Catanzaro	O
server2fi.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Firenze	O
server2mi.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Milano	O
server2rc.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Reggio	O
Calabria	O
server3bo.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Bologna	O
server3ct.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Catania	O
server3.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
server3fi.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Firenze	O
server4fi.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Firenze	O
serverrt.exodus.connexxa	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
Public	O
Resume	O
Confirms	O
Development	O
of	O
Android	B-System
Agent	O
Additionally	O
,	O
an	O
employee	O
of	O
eSurv	B-Organization
quite	O
precisely	O
described	O
their	O
work	O
in	O
developing	O
an	O
"	O
agent	O
to	O
gather	O
data	O
from	O
Android	B-System
devices	O
and	O
send	O
it	O
to	O
a	O
C	O
&	O
C	O
server	O
''	O
as	O
well	O
as	O
researching	O
"	O
vulnerabilities	O
in	O
mobile	O
devices	O
(	O
mainly	O
Android	B-System
)	O
''	O
in	O
a	O
publicly	O
available	O
resume	O
.	O

Further	O
details	O
in	O
it	O
reflect	O
characteristics	O
of	O
Exodus	B-Malware
(	O
such	O
as	O
the	O
bypass	O
of	O
power	O
managers	O
we	O
described	O
from	O
Exodus	B-Malware
One	I-Malware
,	O
and	O
more	O
)	O
:	O
Indicators	O
of	O
Compromise	O
Exodus	B-Malware
One	I-Malware
011b6bcebd543d4eb227e840f04e188fb01f2335b0b81684b60e6b45388d3820	B-Indicator
0f5f1409b1ebbee4aa837d20479732e11399d37f05b47b5359dc53a4001314e5	B-Indicator
2055584625d24687bd027a63bc0b8faa7d1a854a535de74afba24840a52b1d2f	B-Indicator

26fef238028ee4b5b8da631c77bfb44ada3d5db8129c45dea5df6a51c9ea5f55	B-Indicator
33a9da16d096426c82f150e39fc4f9172677885cfeaedcff10c86414e88be802	B-Indicator
34d000ee1e36efd10eb37e2b79d69249d5a85682a61390a89a1b9391c46bf2ba	B-Indicator
4f6146956b50ae3a6e80a1c1f771dba848ba677064eb0e166df5804ac2766898	B-Indicator

5db49122d866967295874ab2c1ce23a7cde50212ff044bbea1da9b49bb9bc149	B-Indicator
70e2eea5609c6954c61f2e5e0a3aea832d0643df93d18d7d78b6f9444dcceef0	B-Indicator
80810a8ec9624f317f832ac2e212dba033212258285344661e5da11b0d9f0b62	B-Indicator
8453ce501fee1ca8a321f16b09969c517f92a24b058ac5b54549eabd58bf1884	B-Indicator

a37f5d2418c5f2f64d06ba28fe62edee1293a56158ddfa9f04020e316054363f	B-Indicator
db59407f72666526fca23d31e3b4c5df86f25eff178e17221219216c6975c63f	B-Indicator
e0acbb0d7e55fb67e550a6bf5cf5c499a9960eaf5f037b785f9004585202593b	B-Indicator
Exodus	B-Malware
One	I-Malware
Package	O
Names	O
com.phonecarrier.linecheck	B-Indicator

rm.rf	I-Indicator
operatore.italia	B-Indicator
it.offertetelefonicheperte	I-Indicator
it.servizipremium	B-Indicator
assistenza.sim	B-Indicator
assistenza.linea.riattiva	B-Indicator
assistenza.linea	B-Indicator
it.promofferte	B-Indicator
Exodus	B-Malware
Two	I-Malware
64c11fdb317d6b7c9930e639f55863df592f23f3c7c861ddd97048891a90c64b	B-Indicator
a42a05bf9b412cd84ea92b166d790e8e72f1d01764f93b05ace62237fbabe40e	B-Indicator
Exodus	B-Malware
Two	I-Malware

ELF	O
Utilities	O
00c787c0c0bc26caf623e66373a5aaa1b913b9caee1f34580bdfdd21954b7cc4	B-Indicator
11499ff2418f4523344de81a447f6786fdba4982057d4114f64db929990b4b59	B-Indicator
13ec6cec511297ac3137cf7d6e4a7c4f5dd2b24478a06262a44f13a3d61070b6	B-Indicator
3c9f08b3280851f54414dfa5a57f40d3b7be7b73736fa0ba21b078e75ce54d33	B-Indicator

3ee3a973c62ba5bd9eab595a7c94b7a26827c5fa5b21964d511ab58903929ec5	B-Indicator
47449a612697ad99a6fbd6e02a84e957557371151f2b034a411ebb10496648c8	B-Indicator
48a7dd672931e408662d2b5e1abcd6ef00097b8ffe3814f0d2799dd6fd74bd88	B-Indicator
824ad333320cbb7873dc49e61c14f749b0e0d88723635524463f2e6f56ea133a	B-Indicator

b46f282f9a1bce3798faee3212e28924730a657eb93cda3824c449868b6ee2e7	B-Indicator
c228a534535b22a316a97908595a2d793d0fecabadc32846c6d1bfb08ca9a658	B-Indicator
e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855	B-Indicator
e3f65f84dd6c2c3a5a653a3788d78920c0321526062a6b53daaf23fa57778a5f	B-Indicator

Command	O
&	O
Controls	O
ad1.fbsba	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
ws.my-local-weather	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
54.71.249	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
137	I-Indicator
54.69.156	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
31	I-Indicator
162.243.172	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
208	I-Indicator
attiva.exodus.esurv	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
it	I-Indicator
The	O
rise	O
of	O
mobile	O
banker	O
Asacub	B-Malware
28	O
AUG	O
2018	O
We	O
encountered	O
the	O
Trojan-Banker.AndroidOS.Asacub	B-Malware
family	O
for	O
the	O
first	O
time	O
in	O
2015	O
,	O
when	O
the	O
first	O
versions	O
of	O
the	O
malware	O
were	O
detected	O
,	O
analyzed	O
,	O
and	O
found	O
to	O
be	O
more	O
adept	O
at	O
spying	O
than	O
stealing	O
funds	O
.	O

The	O
Trojan	O
has	O
evolved	O
since	O
then	O
,	O
aided	O
by	O
a	O
large-scale	O
distribution	O
campaign	O
by	O
its	O
creators	O
(	O
in	O
spring-summer	O
2017	O
)	O
,	O
helping	O
Asacub	B-Malware
to	O
claim	O
top	O
spots	O
in	O
last	O
year	O
’	O
s	O
ranking	O
by	O
number	O
of	O
attacks	O
among	O
mobile	O
banking	O
Trojans	O
,	O
outperforming	O
other	O
families	O
such	O
as	O
Svpeng	B-Malware
and	O
Faketoken	B-Malware
.	O

We	O
decided	O
to	O
take	O
a	O
peek	O
under	O
the	O
hood	O
of	O
a	O
modern	O
member	O
of	O
the	O
Asacub	B-Malware
family	O
.	O

Our	O
eyes	O
fell	O
on	O
the	O
latest	O
version	O
of	O
the	O
Trojan	O
,	O
which	O
is	O
designed	O
to	O
steal	O
money	O
from	O
owners	O
of	O
Android	B-System
devices	O
connected	O
to	O
the	O
mobile	O
banking	O
service	O
of	O
one	O
of	O
Russia	O
’	O
s	O
largest	O
banks	O
.	O

Asacub	B-Malware
versions	O
Sewn	O
into	O
the	O
body	O
of	O
the	O
Trojan	O
is	O
the	O
version	O
number	O
,	O
consisting	O
of	O
two	O
or	O
three	O
digits	O
separated	O
by	O
periods	O
.	O

The	O
numbering	O
seems	O
to	O
have	O
started	O
anew	O
after	O
the	O
version	O
9	O
.	O

The	O
name	O
Asacub	B-Malware
appeared	O
with	O
version	O
4	O
in	O
late	O
2015	O
;	O
previous	O
versions	O
were	O
known	O
as	O
Trojan-SMS.AndroidOS.Smaps	B-Indicator
.	O

Versions	O
5.X.X-8.X.X	O
were	O
active	O
in	O
2016	O
,	O
and	O
versions	O
9.X.X-1.X.X	O
in	O
2017	O
.	O

In	O
2018	O
,	O
the	O
most	O
actively	O
distributed	O
versions	O
were	O
5.0.0	O
and	O
5.0.3	O
.	O

Communication	O
with	O
C	O
&	O
C	O
Although	O
Asacub	B-Malware
’	O
s	O
capabilities	O
gradually	O
evolved	O
,	O
its	O
network	O
behavior	O
and	O
method	O
of	O
communication	O
with	O
the	O
command-and-control	O
(	O
C	O
&	O
C	O
)	O
server	O
changed	O
little	O
.	O

This	O
strongly	O
suggested	O
that	O
the	O
banking	O
Trojans	O
,	O
despite	O
differing	O
in	O
terms	O
of	O
capability	O
,	O
belong	O
to	O
the	O
same	O
family	O
.	O

Data	O
was	O
always	O
sent	O
to	O
the	O
C	O
&	O
C	O
server	O
via	O
HTTP	O
in	O
the	O
body	O
of	O
a	O
POST	O
request	O
in	O
encrypted	O
form	O
to	O
the	O
relative	O
address	O
/something/index.php	B-Indicator
.	O

In	O
earlier	O
versions	O
,	O
the	O
something	O
part	O
of	O
the	O
relative	O
path	O
was	O
a	O
partially	O
intelligible	O
,	O
yet	O
random	O
mix	O
of	O
words	O
and	O
short	O
combinations	O
of	O
letters	O
and	O
numbers	O
separated	O
by	O
an	O
underscore	O
,	O
for	O
example	O
,	O
“	O
bee_bomb	O
”	O
or	O
“	O
my_te2_mms	O
”	O
.	O

Example	O
of	O
traffic	O
from	O
an	O
early	O
version	O
of	O
Asacub	B-Malware
(	O
2015	O
)	O
The	O
data	O
transmitted	O
and	O
received	O
is	O
encrypted	O
with	O
the	O
RC4	O
algorithm	O
and	O
encoded	O
using	O
the	O
base64	O
standard	O
.	O

The	O
C	O
&	O
C	O
address	O
and	O
the	O
encryption	O
key	O
(	O
one	O
for	O
different	O
modifications	O
in	O
versions	O
4.x	O
and	O
5.x	O
,	O
and	O
distinct	O
for	O
different	O
C	O
&	O
Cs	O
in	O
later	O
versions	O
)	O
are	O
stitched	O
into	O
the	O
body	O
of	O
the	O
Trojan	O
.	O

In	O
early	O
versions	O
of	O
Asacub	B-Malware
,	O
.com	O
,	O
.biz	O
,	O
.info	O
,	O
.in	O
,	O
.pw	O
were	O
used	O
as	O
top-level	O
domains	O
.	O

In	O
the	O
2016	O
version	O
,	O
the	O
value	O
of	O
the	O
User-Agent	O
header	O
changed	O
,	O
as	O
did	O
the	O
method	O
of	O
generating	O
the	O
relative	O
path	O
in	O
the	O
URL	O
:	O
now	O
the	O
part	O
before	O
/index.php	O
is	O
a	O
mix	O
of	O
a	O
pronounceable	O
(	O
if	O
not	O
entirely	O
meaningful	O
)	O
word	O
and	O
random	O
letters	O
and	O
numbers	O
,	O
for	O
example	O
,	O
“	O
muromec280j9tqeyjy5sm1qy71	B-Indicator
”	O
or	O
“	O
parabbelumf8jgybdd6w0qa0	B-Indicator
”	O
.	O

Moreover	O
,	O
incoming	O
traffic	O
from	O
the	O
C	O
&	O
C	O
server	O
began	O
to	O
use	O
gzip	O
compression	O
,	O
and	O
the	O
top-level	O
domain	O
for	O
all	O
C	O
&	O
Cs	O
was	O
.com	O
:	O
Since	O
December	O
2016	O
,	O
the	O
changes	O
in	O
C	O
&	O
C	O
communication	O
methods	O
have	O
affected	O
only	O
how	O
the	O
relative	O
path	O
in	O
the	O
URL	O
is	O
generated	O
:	O
the	O
pronounceable	O
word	O
was	O
replaced	O
by	O
a	O
rather	O
long	O
random	O
combination	O
of	O
letters	O
and	O
numbers	O
,	O
for	O
example	O
,	O
“	O
ozvi4malen7dwdh	B-Indicator
”	O
or	O
“	O
f29u8oi77024clufhw1u5ws62	B-Indicator
”	O
.	O

At	O
the	O
time	O
of	O
writing	O
this	O
article	O
,	O
no	O
other	O
significant	O
changes	O
in	O
Asacub	B-Malware
’	O
s	O
network	O
behavior	O
had	O
been	O
observed	O
:	O
The	O
origin	O
of	O
Asacub	B-Malware
It	O
is	O
fairly	O
safe	O
to	O
say	O
that	O
the	O
Asacub	B-Malware
family	O
evolved	O
from	O
Trojan-SMS.AndroidOS.Smaps	B-Indicator
.	O

Communication	O
between	O
both	O
Trojans	O
and	O
their	O
C	O
&	O
C	O
servers	O
is	O
based	O
on	O
the	O
same	O
principle	O
,	O
the	O
relative	O
addresses	O
to	O
which	O
Trojans	O
send	O
network	O
requests	O
are	O
generated	O
in	O
a	O
similar	O
manner	O
,	O
and	O
the	O
set	O
of	O
possible	O
commands	O
that	O
the	O
two	O
Trojans	O
can	O
perform	O
also	O
overlaps	O
.	O

What	O
’	O
s	O
more	O
,	O
the	O
numbering	O
of	O
Asacub	B-Malware
versions	O
is	O
a	O
continuation	O
of	O
the	O
Smaps	B-Malware
system	O
.	O

The	O
main	O
difference	O
is	O
that	O
Smaps	B-Malware
transmits	O
data	O
as	O
plain	O
text	O
,	O
while	O
Asacub	B-Malware
encrypts	O
data	O
with	O
the	O
RC4	O
algorithm	O
and	O
then	O
encodes	O
it	O
into	O
base64	O
format	O
.	O

Let	O
’	O
s	O
compare	O
examples	O
of	O
traffic	O
from	O
Smaps	B-Malware
and	O
Asacub	B-Malware
—	O
an	O
initializing	O
request	O
to	O
the	O
C	O
&	O
C	O
server	O
with	O
information	O
about	O
the	O
infected	O
device	O
and	O
a	O
response	O
from	O
the	O
server	O
with	O
a	O
command	O
for	O
execution	O
:	O
Smaps	B-Malware
request	O
Asacub	B-Malware
request	O
Decrypted	O
data	O
from	O
Asacub	B-Malware
traffic	O
:	O
{	O
“	O
id	O
”	O
:	O
”	O
532bf15a-b784-47e5-92fa-72198a2929f5″	B-Indicator
,	O
”	O
type	O
”	O
:	O
”	O
get	O
”	O
,	O
”	O
info	O
”	O
:	O
”	O
imei:365548770159066	O
,	O
country	O
:	O
PL	O
,	O
cell	O
:	O
Tele2	O

,	O
android:4.2.2	O
,	O
model	O
:	O
GT-N5100	O
,	O
phonenumber	O
:	O
+486679225120	O
,	O
sim:6337076348906359089f	O
,	O
app	O
:	O
null	O
,	O
ver:5.0.2″	O
}	O
Data	O
sent	O
to	O
the	O
server	O
[	O
{	O
“	O
command	O
”	O
:	O
”	O
sent	O
&	O
&	O
&	O
”	O
,	O
”	O
params	O
”	O
:	O
{	O
“	O
to	O
”	O
:	O
”	O
+79262000900″	O
,	O
”	O
body	O
”	O
:	O
”	O
\u0410\u0412\u0422\u041e\u041f\u041b\u0410\u0422\u0415\u0416	O

1000	O
50″	O
,	O
”	O
timestamp	O
”	O
:	O
”	O
1452272572″	O
}	O
}	O
,	O
{	O
“	O
command	O
”	O
:	O
”	O
sent	O
&	O
&	O
&	O
”	O
,	O
”	O
params	O
”	O
:	O
{	O
“	O
to	O
”	O
:	O
”	O
+79262000900″	O
,	O
”	O
body	O
”	O
:	O
”	O
BALANCE	O
”	O
,	O
”	O
timestamp	O
”	O
:	O
”	O
1452272573″	O
}	O
}	O
]	O
Instructions	O
received	O
from	O
the	O
server	O
A	O
comparison	O
can	O
also	O
be	O
made	O
of	O
the	O
format	O
in	O
which	O
Asacub	O
and	O
Smaps	O
forward	O
incoming	O
SMS	O
(	O
encoded	O
with	O
the	O
base64	O
algorithm	O
)	O
from	O
the	O
device	O
to	O
the	O
C	O
&	O
C	O
server	O
:	O
Smaps	O

format	O
Asacub	O
format	O
Decrypted	O
data	O
from	O
Asacub	O
traffic	O
:	O
{	O
“	O
data	O
”	O
:	O
”	O
2015:10:14_02:41:15″	O
,	O
”	O
id	O
”	O
:	O
”	O
532bf15a-b784-47e5-92fa-72198a2929f5″	B-Indicator
,	O
”	O
text	O
”	O
:	O
”	O
SSB0aG91Z2h0IHdlIGdvdCBwYXN0IHRoaXMhISBJJ20gbm90IGh1bmdyeSBhbmQgbmU=	O
”	O
,	O
”	O
number	O
”	O
:	O
”	O
1790″	O
,	O
”	O
type	O

”	O
:	O
”	O
load	O
”	O
}	O
Propagation	O
The	O
banking	O
Trojan	O
is	O
propagated	O
via	O
phishing	O
SMS	O
containing	O
a	O
link	O
and	O
an	O
offer	O
to	O
view	O
a	O
photo	O
or	O
MMS	O
.	O

The	O
link	O
points	O
to	O
a	O
web	O
page	O
with	O
a	O
similar	O
sentence	O
and	O
a	O
button	O
for	O
downloading	O
the	O
APK	O
file	O
of	O
the	O
Trojan	O
to	O
the	O
device	O
.	O

The	O
Trojan	O
download	O
window	O
Asacub	B-Malware
masquerades	O
under	O
the	O
guise	O
of	O
an	O
MMS	O
app	O
or	O
a	O
client	O
of	O
a	O
popular	O
free	O
ads	O
service	O
.	O

We	O
came	O
across	O
the	O
names	O
Photo	O
,	O
Message	O
,	O
Avito	O
Offer	O
,	O
and	O
MMS	O
Message	O
.	O

App	O
icons	O
under	O
which	O
Asacub	B-Malware
masks	O
itself	O
The	O
APK	O
files	O
of	O
the	O
Trojan	O
are	O
downloaded	O
from	O
sites	O
such	O
as	O
mmsprivate	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
site	I-Indicator
,	O
photolike	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
fun	I-Indicator
,	O
you-foto	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
site	I-Indicator
,	O
and	O
mms4you	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
me	I-Indicator
under	O
names	O
in	O
the	O
format	O
:	O
photo_	B-Indicator
[	I-Indicator
number	I-Indicator
]	I-Indicator
_img.apk	I-Indicator
,	O
mms_	B-Indicator
[	I-Indicator
number	I-Indicator
]	I-Indicator
_img.apk	I-Indicator
avito_	B-Indicator
[	I-Indicator
number	I-Indicator
]	I-Indicator
.apk	I-Indicator
,	O
mms.img_	B-Indicator
[	I-Indicator
number	I-Indicator
]	I-Indicator
_photo.apk	I-Indicator
,	O
mms	B-Indicator
[	I-Indicator
number	I-Indicator
]	I-Indicator
_photo.image.apk	I-Indicator
,	O
mms	B-Indicator
[	I-Indicator
number	I-Indicator
]	I-Indicator
_photo.img.apk	I-Indicator
,	O
mms.img.photo_	B-Indicator
[	I-Indicator
number	I-Indicator
]	I-Indicator
.apk	I-Indicator
,	O
photo_	B-Indicator
[	I-Indicator
number	I-Indicator
]	I-Indicator
_obmen.img.apk	I-Indicator
.	O

For	O
the	O
Trojan	O
to	O
install	O
,	O
the	O
user	O
must	O
allow	O
installation	O
of	O
apps	O
from	O
unknown	O
sources	O
in	O
the	O
device	O
settings	O
.	O

Infection	O
During	O
installation	O
,	O
depending	O
on	O
the	O
version	O
of	O
the	O
Trojan	O
,	O
Asacub	B-Malware
prompts	O
the	O
user	O
either	O
for	O
Device	O
Administrator	O
rights	O
or	O
for	O
permission	O
to	O
use	O
AccessibilityService	O
.	O

After	O
receiving	O
the	O
rights	O
,	O
it	O
sets	O
itself	O
as	O
the	O
default	O
SMS	O
app	O
and	O
disappears	O
from	O
the	O
device	O
screen	O
.	O

If	O
the	O
user	O
ignores	O
or	O
rejects	O
the	O
request	O
,	O
the	O
window	O
reopens	O
every	O
few	O
seconds	O
.	O

The	O
Trojan	O
requests	O
Device	O
Administrator	O
rights	O
The	O
Trojan	O
requests	O
permission	O
to	O
use	O
AccessibilityService	O
After	O
installation	O
,	O
the	O
Trojan	O
starts	O
communicating	O
with	O
the	O
cybercriminals	O
’	O
C	O
&	O
C	O
server	O
.	O

All	O
data	O
is	O
transmitted	O
in	O
JSON	O
format	O
(	O
after	O
decryption	O
)	O
.	O

It	O
includes	O
information	O
about	O
the	O
smartphone	O
model	O
,	O
the	O
OS	O
version	O
,	O
the	O
mobile	O
operator	O
,	O
and	O
the	O
Trojan	O
version	O
.	O

Let	O
’	O
s	O
take	O
an	O
in-depth	O
look	O
at	O
Asacub	B-Malware
5.0.3	O
,	O
the	O
most	O
widespread	O
version	O
in	O
2018	O
.	O

Structure	O
of	O
data	O
sent	O
to	O
the	O
server	O
:	O
To	O
begin	O
with	O
,	O
the	O
Trojan	O
sends	O
information	O
about	O
the	O
device	O
to	O
the	O
server	O
:	O
In	O
response	O
,	O
the	O
server	O
sends	O
the	O
code	O
of	O
the	O
command	O
for	O
execution	O
(	O
“	O
command	O
”	O
)	O
,	O
its	O
parameters	O
(	O
“	O
params	O
”	O
)	O
,	O
and	O
the	O
time	O
delay	O
before	O
execution	O
(	O
“	O
waitrun	O
”	O
in	O
milliseconds	O
)	O
.	O

List	O
of	O
commands	O
sewn	O
into	O
the	O
body	O
of	O
the	O
Trojan	O
:	O
Command	O
code	O
Parameters	O
Actions	O
2	O
–	O
Sending	O
a	O
list	O
of	O
contacts	O
from	O
the	O
address	B-System
book	I-System
of	O
the	O
infected	O
device	O
to	O
the	O
C	O
&	O
C	O
server	O
7	O
“	O
to	O
”	O
:	O
int	O
Calling	O
the	O
specified	O
number	O
11	O
“	O
to	O
”	O
:	O
int	O
,	O
“	O
body	O
”	O
:	O
string	O
Sending	O
an	O
SMS	O
with	O
the	O
specified	O
text	O
to	O
the	O
specified	O
number	O
19	O
“	O
text	O
”	O
:	O
string	O
,	O
“	O
n	O
”	O
:	O
string	O
Sending	O
SMS	O
with	O
the	O
specified	O
text	O
to	O
numbers	O
from	O
the	O
address	B-System
book	I-System
of	O
the	O
infected	O
device	O
,	O
with	O
the	O
name	O
of	O
the	O
addressee	O
from	O
the	O

address	B-System
book	I-System
substituted	O
into	O
the	O
message	O
text	O
40	O
“	O
text	O
”	O
:	O
string	O
Shutting	O
down	O
applications	O
with	O
specific	O
names	O
(	O
antivirus	O
and	O
banking	O
applications	O
)	O
The	O
set	O
of	O
possible	O
commands	O
is	O
the	O
most	O
significant	O
difference	O
between	O
the	O
various	O
flavors	O
of	O
Asacub	B-Malware
.	O

In	O
the	O
2015-early	O
2016	O
versions	O
examined	O
in	O
this	O
article	O
,	O
C	O
&	O
C	O
instructions	O
in	O
JSON	O
format	O
contained	O
the	O
name	O
of	O
the	O
command	O
in	O
text	O
form	O
(	O
“	O
get_sms	O
”	O
,	O
“	O
block_phone	O
”	O
)	O
.	O

In	O
later	O
versions	O
,	O
instead	O
of	O
the	O
name	O
of	O
the	O
command	O
,	O
its	O
numerical	O
code	O
was	O
transmitted	O
.	O

The	O
same	O
numerical	O
code	O
corresponded	O
to	O
one	O
command	O
in	O
different	O
versions	O
,	O
but	O
the	O
set	O
of	O
supported	O
commands	O
varied	O
.	O

For	O
example	O
,	O
version	O
9.0.7	O
(	O
2017	O
)	O
featured	O
the	O
following	O
set	O
of	O
commands	O
:	O
2	O
,	O
4	O
,	O
8	O
,	O
11	O
,	O
12	O
,	O
15	O
,	O
16	O
,	O
17	O
,	O
18	O
,	O
19	O
,	O
20	O
.	O

After	O
receiving	O
the	O
command	O
,	O
the	O
Trojan	O
attempts	O
to	O
execute	O
it	O
,	O
before	O
informing	O
C	O
&	O
C	O
of	O
the	O
execution	O
status	O
and	O
any	O
data	O
received	O
.	O

The	O
“	O
id	O
”	O
value	O
inside	O
the	O
“	O
data	O
”	O
block	O
is	O
equal	O
to	O
the	O
“	O
timestamp	O
”	O
value	O
of	O
the	O
relevant	O
command	O
:	O
In	O
addition	O
,	O
the	O
Trojan	O
sets	O
itself	O
as	O
the	O
default	O
SMS	O
application	O
and	O
,	O
on	O
receiving	O
a	O
new	O
SMS	O
,	O
forwards	O
the	O
sender	O
’	O
s	O
number	O
and	O
the	O
message	O
text	O
in	O
base64	O
format	O
to	O
the	O
cybercriminal	O
:	O
Thus	O
,	O
Asacub	B-Malware
can	O
withdraw	O
funds	O
from	O
a	O
bank	O
card	O
linked	O
to	O
the	O
phone	O
by	O
sending	O
SMS	O
for	O
the	O
transfer	O
of	O
funds	O
to	O
another	O
account	O
using	O
the	O
number	O
of	O
the	O
card	O
or	O
mobile	O
phone	O
.	O

Moreover	O
,	O
the	O
Trojan	O
intercepts	O
SMS	O
from	O
the	O
bank	O
that	O
contain	O
one-time	O
passwords	O
and	O
information	O
about	O
the	O
balance	O
of	O
the	O
linked	O
bank	O
card	O
.	O

Some	O
versions	O
of	O
the	O
Trojan	O
can	O
autonomously	O
retrieve	O
confirmation	O
codes	O
from	O
such	O
SMS	O
and	O
send	O
them	O
to	O
the	O
required	O
number	O
.	O

What	O
’	O
s	O
more	O
,	O
the	O
user	O
can	O
not	O
check	O
the	O
balance	O
via	O
mobile	O
banking	O
or	O
change	O
any	O
settings	O
there	O
,	O
because	O
after	O
receiving	O
the	O
command	O
with	O
code	O
40	O
,	O
the	O
Trojan	O
prevents	O
the	O
banking	O
app	O
from	O
running	O
on	O
the	O
phone	O
.	O

User	O
messages	O
created	O
by	O
the	O
Trojan	O
during	O
installation	O
typically	O
contain	O
grammatical	O
and	O
spelling	O
errors	O
,	O
and	O
use	O
a	O
mixture	O
of	O
Cyrillic	O
and	O
Latin	O
characters	O
.	O

The	O
Trojan	O
also	O
employs	O
various	O
obfuscation	O
methods	O
:	O
from	O
the	O
simplest	O
,	O
such	O
as	O
string	O
concatenation	O
and	O
renaming	O
of	O
classes	O
and	O
methods	O
,	O
to	O
implementing	O
functions	O
in	O
native	O
code	O
and	O
embedding	O
SO	O
libraries	O
in	O
C/C++	O
in	O
the	O
APK	O
file	O
,	O
which	O
requires	O
the	O
use	O
of	O
additional	O
tools	O
or	O
dynamic	O
analysis	O
for	O
deobfuscation	O
,	O
since	O
most	O
tools	O
for	O
static	O
analysis	O
of	O
Android	O
apps	O
support	O
only	O
Dalvik	O
bytecode	O
.	O

In	O
some	O
versions	O
of	O
Asacub	B-Malware
,	O
strings	O
in	O
the	O
app	O
are	O
encrypted	O
using	O
the	O
same	O
algorithm	O
as	O
data	O
sent	O
to	O
C	O
&	O
C	O
,	O
but	O
with	O
different	O
keys	O
.	O

Example	O
of	O
using	O
native	O
code	O
for	O
obfuscation	O
Examples	O
of	O
using	O
string	O
concatenation	O
for	O
obfuscation	O
Example	O
of	O
encrypting	O
strings	O
in	O
the	O
Trojan	O
Asacub	B-Malware
distribution	O
geography	O
Asacub	B-Malware
is	O
primarily	O
aimed	O
at	O
Russian	O
users	O
:	O
98	O
%	O
of	O
infections	O
(	O
225,000	O
)	O
occur	O
in	O
Russia	O
,	O
since	O
the	O
cybercriminals	O
specifically	O
target	O
clients	O
of	O
a	O
major	O
Russian	O
bank	O
.	O

The	O
Trojan	O
also	O
hit	O
users	O
from	O
Ukraine	O
,	O
Turkey	O
,	O
Germany	O
,	O
Belarus	O
,	O
Poland	O
,	O
Armenia	O
,	O
Kazakhstan	O
,	O
the	O
US	O
,	O
and	O
other	O
countries	O
.	O

Conclusion	O
The	O
case	O
of	O
Asacub	B-Malware
shows	O
that	O
mobile	O
malware	O
can	O
function	O
for	O
several	O
years	O
with	O
minimal	O
changes	O
to	O
the	O
distribution	O
scheme	O
.	O

It	O
is	O
basically	O
SMS	O
spam	O
:	O
many	O
people	O
still	O
follow	O
suspicious	O
links	O
,	O
install	O
software	O
from	O
third-party	O
sources	O
,	O
and	O
give	O
permissions	O
to	O
apps	O
without	O
a	O
second	O
thought	O
.	O

At	O
the	O
same	O
time	O
,	O
cybercriminals	O
are	O
reluctant	O
to	O
change	O
the	O
method	O
of	O
communication	O
with	O
the	O
C	O
&	O
C	O
server	O
,	O
since	O
this	O
would	O
require	O
more	O
effort	O
and	O
reap	O
less	O
benefit	O
than	O
modifying	O
the	O
executable	O
file	O
.	O

The	O
most	O
significant	O
change	O
in	O
this	O
particular	O
Trojan	O
’	O
s	O
history	O
was	O
the	O
encryption	O
of	O
data	O
sent	O
between	O
the	O
device	O
and	O
C	O
&	O
C	O
.	O

That	O
said	O
,	O
so	O
as	O
to	O
hinder	O
detection	O
of	O
new	O
versions	O
,	O
the	O
Trojan	O
’	O
s	O
APK	O
file	O
and	O
the	O
C	O
&	O
C	O
server	O
domains	O
are	O
changed	O
regularly	O
,	O
and	O
the	O
Trojan	O
download	O
links	O
are	O
often	O
one-time-use	O
.	O

IOCs	O
C	O
&	O
C	O
IP	O
addresses	O
:	O
155.133.82.181	B-Indicator
155.133.82.240	B-Indicator
155.133.82.244	B-Indicator
185.234.218.59	B-Indicator
195.22.126.160	B-Indicator
195.22.126.163	B-Indicator
195.22.126.80	B-Indicator
195.22.126.81	B-Indicator
5.45.73.24	B-Indicator
5.45.74.130	B-Indicator
IP	I-Indicator
addresses	I-Indicator
from	O
which	O
the	O
Trojan	O
was	O
downloaded	O
:	O
185.174.173.31	B-Indicator
185.234.218.59	B-Indicator
188.166.156.110	B-Indicator
195.22.126.160	B-Indicator
195.22.126.80	B-Indicator
195.22.126.81	B-Indicator

195.22.126.82	B-Indicator
195.22.126.83	B-Indicator
SHA256	O
:	O
158c7688877853ffedb572ccaa8aa9eff47fa379338151f486e46d8983ce1b67	B-Indicator
3aedbe7057130cf359b9b57fa533c2b85bab9612c34697585497734530e7457d	B-Indicator
f3ae6762df3f2c56b3fe598a9e3ff96ddf878c553be95bacbd192bd14debd637	B-Indicator
df61a75b7cfa128d4912e5cb648cfc504a8e7b25f6c83ed19194905fef8624c8	B-Indicator

c0cfd462ab21f6798e962515ac0c15a92036edd3e2e63639263bf2fd2a10c184	B-Indicator
d791e0ce494104e2ae0092bb4adc398ce740fef28fa2280840ae7f61d4734514	B-Indicator
38dcec47e2f4471b032a8872ca695044ddf0c61b9e8d37274147158f689d65b9	B-Indicator
27cea60e23b0f62b4b131da29fdda916bc4539c34bb142fb6d3f8bb82380fe4c	B-Indicator

31edacd064debdae892ab0bc788091c58a03808997e11b6c46a6a5de493ed25d	B-Indicator
87ffec0fe0e7a83e6433694d7f24cfde2f70fc45800aa2acb8e816ceba428951	B-Indicator
eabc604fe6b5943187c12b8635755c303c450f718cc0c8e561df22a27264f101	B-Indicator
Kernel	O
Backdoor	O
found	O
in	O
Gadgets	O
Powered	O
by	O
Popular	O
Chinese	O
ARM	B-System

Maker	O
May	O
12	O
,	O
2016	O
Mohit	O
Kumar	O
How	O
to	O
Hack	O
an	O
Android	B-System
device	O
?	O

It	O
is	O
possibly	O
one	O
of	O
the	O
most	O
frequently	O
asked	O
questions	O
on	O
the	O
Internet	O
.	O

Although	O
it	O
's	O
not	O
pretty	O
simple	O
to	O
hack	O
Android	B-System
devices	O
and	O
gadgets	O
,	O
sometimes	O
you	O
just	O
get	O
lucky	O
to	O
find	O
a	O
backdoor	O
access	O
.	O

Thanks	O
to	O
Allwinner	B-Organization
,	O
a	O
Chinese	O
ARM	B-System
system-on-a-chip	O
maker	O
,	O
which	O
has	O
recently	O
been	O
caught	O
shipping	O
a	O
version	O
of	O
Linux	B-System
Kernel	O
with	O
an	O
incredibly	O
simple	O
and	O
easy-to-use	O
built-in	O
backdoor	O
.	O

Chinese	O
fabless	O
semiconductor	O
company	O
Allwinner	B-Organization
is	O
a	O
leading	O
supplier	O
of	O
application	O
processors	O
that	O
are	O
used	O
in	O
many	O
low-cost	O
Android	B-System
tablets	O
,	O
ARM-based	B-Organization
PCs	O
,	O
set-top	O
boxes	O
,	O
and	O
other	O
electronic	O
devices	O
worldwide	O
.	O

Simple	O
Backdoor	O
Exploit	O
to	O
Hack	O
Android	B-System
Devices	O
All	O
you	O
need	O
to	O
do	O
to	O
gain	O
root	O
access	O
of	O
an	O
affected	O
Android	B-System
device	O
is…	O
Send	O
the	O
text	O
"	O
rootmydevice	O
''	O
to	O
any	O
undocumented	O
debugging	O
process	O
.	O

The	O
local	O
privileges	O
escalation	O
backdoor	O
code	O
for	O
debugging	O
ARM-powered	B-System
Android	B-System
devices	O
managed	O
to	O
make	O
its	O
way	O
in	O
shipped	O
firmware	O
after	O
firmware	O
makers	O
wrote	O
their	O
own	O
kernel	O
code	O
underneath	O
a	O
custom	O
Android	B-System
build	O
for	O
their	O
devices	O
,	O
though	O
the	O
mainstream	O
kernel	O
source	O
is	O
unaffected	O
.	O

The	O
backdoor	O
code	O
is	O
believed	O
to	O
have	O
been	O
left	O
by	O
mistake	O
by	O
the	O
authors	O
after	O
completing	O
the	O
debugging	O
process	O
.	O

For	O
exploiting	O
this	O
issue	O
,	O
any	O
process	O
running	O
with	O
any	O
UID	O
can	O
be	O
converted	O
into	O
root	O
easily	O
by	O
simply	O
using	O
the	O
following	O
command	O
:	O
echo	O
"	O
rootmydevice	B-Indicator
''	O
>	I-Indicator
/proc/sunxi_debug/sunxi_debug	I-Indicator
The	O
Linux	B-Indicator
3.4-sunxi	I-Indicator
kernel	O
was	O
originally	O
designed	O
to	O
support	O
the	O
Android	B-System
operating	O
system	O
on	O
Allwinner	B-Organization
ARM	B-System
for	O
tablets	O
,	O
but	O
later	O
it	O
was	O
used	O
to	O
port	O
Linux	B-System
to	O
many	O
Allwinner	B-Organization
processors	O
on	O
boards	O
like	O
Banana	B-System
Pi	I-System
micro-PCs	I-System
,	O
Orange	B-System
Pi	I-System
,	O
and	O
other	O
devices	O
.	O

At	O
the	O
forum	O
of	O
the	O
Armbian	B-System
operating	O
system	O
,	O
a	O
moderator	O
who	O
goes	O
by	O
the	O
name	O
Tkaiser	O
noted	O
that	O
the	O
backdoor	O
code	O
could	O
remotely	O
be	O
exploitable	O
"	O
if	O
combined	O
with	O
networked	O
services	O
that	O
might	O
allow	O
access	O
to	O
/proc	B-Indicator
.	O

''	O
This	O
security	O
hole	O
is	O
currently	O
present	O
in	O
every	O
operating	O
system	O
image	O
for	O
A83T	B-System
,	O
H3	B-System
or	O
H8	B-System
devices	O
that	O
rely	O
on	O
kernel	B-System
3.4	I-System
,	O
he	O
added	O
.	O

This	O
blunder	O
made	O
by	O
the	O
company	O
has	O
been	O
frustrating	O
to	O
many	O
developers	O
.	O

Allwinner	B-Organization
has	O
also	O
been	O
less	O
transparent	O
about	O
the	O
backdoor	O
code	O
.	O

David	O
Manouchehri	O
released	O
the	O
information	O
about	O
the	O
backdoor	O
through	O
its	O
own	O
Github	B-Organization
account	O
(	O
Pastebin	B-Organization
)	O
and	O
then	O
apparently	O
deleted	O
it	O
.	O

Mobile	O
Malware	O
Evolution	O
:	O
2013	O
24	O
FEB	O
2014	O
The	O
mobile	O
malware	O
sector	O
is	O
growing	O
rapidly	O
both	O
technologically	O
and	O
structurally	O
.	O

It	O
is	O
safe	O
to	O
say	O
that	O
today	O
’	O
s	O
cybercriminal	O
is	O
no	O
longer	O
a	O
lone	O
hacker	O
but	O
part	O
of	O
a	O
serious	O
business	O
operation	O
.	O

There	O
are	O
various	O
types	O
of	O
actors	O
involved	O
in	O
the	O
mobile	O
malware	O
industry	O
:	O
virus	O
writers	O
,	O
testers	O
,	O
interface	O
designers	O
of	O
both	O
the	O
malicious	O
apps	O
and	O
the	O
web	O
pages	O
they	O
are	O
distributed	O
from	O
,	O
owners	O
of	O
the	O
partner	O
programs	O
that	O
spread	O
the	O
malware	O
,	O
and	O
mobile	O
botnet	O
owners	O
.	O

This	O
division	O
of	O
labor	O
among	O
the	O
cybercriminals	O
can	O
also	O
be	O
seen	O
in	O
the	O
behavior	O
of	O
their	O
Trojans	O
.	O

In	O
2013	O
,	O
there	O
was	O
evidence	O
of	O
cooperation	O
(	O
most	O
probably	O
on	O
a	O
commercial	O
basis	O
)	O
between	O
different	O
groups	O
of	O
virus	O
writers	O
.	O

For	O
example	O
,	O
the	O
botnet	O
Trojan-SMS.AndroidOS.Opfake.a	B-Malware
,	O
in	O
addition	O
to	O
its	O
own	O
activity	O
,	O
also	O
spread	O
Backdoor.AndroidOS.Obad.a	B-Malware
by	O
sending	O
spam	O
containing	O
a	O
link	O
to	O
the	O
malware	O
to	O
the	O
victim	O
’	O
s	O
list	O
of	O
contacts	O
.	O

It	O
is	O
now	O
clear	O
that	O
a	O
distinct	O
industry	O
has	O
developed	O
and	O
is	O
becoming	O
more	O
focused	O
on	O
extracting	O
profits	O
,	O
which	O
is	O
clearly	O
evident	O
from	O
the	O
functionality	O
of	O
the	O
malware	O
.	O

2013	O
in	O
figures	O
A	O
total	O
of	O
143,211	O
new	O
modifications	O
of	O
malicious	O
programs	O
targeting	O
mobile	O
devices	O
were	O
detected	O
in	O
all	O
of	O
2013	O
(	O
as	O
of	O
January	O
1	O
,	O
2014	O
)	O
.	O

In	O
2013	O
,	O
3,905,502	O
installation	O
packages	O
were	O
used	O
by	O
cybercriminals	O
to	O
distribute	O
mobile	O
malware	O
.	O

Overall	O
in	O
2012-2013	O
we	O
detected	O
approximately	O
10,000,000	O
unique	O
malicious	O
installation	O
packages	O
:	O
Different	O
installation	O
packages	O
can	O
install	O
programs	O
with	O
the	O
same	O
functionality	O
that	O
differ	O
only	O
in	O
terms	O
of	O
the	O
malicious	O
app	O
interface	O
and	O
,	O
for	O
instance	O
,	O
the	O
content	O
of	O
the	O
text	O
messages	O
it	O
spreads	O
.	O

Android	B-System
remains	O
a	O
prime	O
target	O
for	O
malicious	O
attacks	O
.	O

98.05	O
%	O
of	O
all	O
malware	O
detected	O
in	O
2013	O
targeted	O
this	O
platform	O
,	O
confirming	O
both	O
the	O
popularity	O
of	O
this	O
mobile	O
OS	O
and	O
the	O
vulnerability	O
of	O
its	O
architecture	O
.	O

Most	O
mobile	O
malware	O
is	O
designed	O
to	O
steal	O
users	O
’	O
money	O
,	O
including	O
SMS-Trojans	O
,	O
and	O
lots	O
of	O
backdoors	O
and	O
Trojans	O
.	O

Over	O
the	O
year	O
,	O
the	O
number	O
of	O
mobile	O
malware	O
modifications	O
designed	O
for	O
phishing	O
,	O
the	O
theft	O
of	O
credit	O
card	O
information	O
and	O
money	O
increased	O
by	O
a	O
factor	O
of	O
19.7	O
.	O

In	O
2013	O
,	O
Kaspersky	B-Organization
Lab	I-Organization
mobile	O
products	O
prevented	O
2,500	O
infections	O
by	O
banking	O
Trojans	O
.	O

Methods	O
and	O
techniques	O
2013	O
not	O
only	O
saw	O
a	O
radical	O
increase	O
in	O
output	O
from	O
mobile	O
virus	O
writers	O
but	O
also	O
saw	O
them	O
actively	O
applying	O
methods	O
and	O
technologies	O
that	O
allowed	O
cybercriminals	O
to	O
use	O
their	O
malware	O
more	O
effectively	O
.	O

There	O
were	O
several	O
distinct	O
areas	O
where	O
mobile	O
malware	O
underwent	O
advances	O
.	O

Distribution	O
Cybercriminals	O
made	O
use	O
of	O
some	O
exceptionally	O
sophisticated	O
methods	O
to	O
infect	O
mobile	O
devices	O
.	O

Infecting	O
legal	O
web	O
resources	O
help	O
spread	O
mobile	O
malware	O
via	O
popular	O
websites	O
.	O

More	O
and	O
more	O
smartphone	O
and	O
tablet	O
owners	O
use	O
their	O
devices	O
to	O
access	O
websites	O
,	O
unaware	O
that	O
even	O
the	O
most	O
reputable	O
resources	O
can	O
be	O
hacked	O
.	O

According	O
to	O
our	O
data	O
,	O
0.4	O
%	O
of	O
the	O
websites	O
visited	O
by	O
users	O
of	O
our	O
products	O
were	O
compromised	O
sites	O
.	O

Distribution	O
via	O
alternative	O
app	O
stores	O
.	O

In	O
Asia	O
there	O
are	O
numerous	O
companies	O
producing	O
Android-based	B-System
devices	O
and	O
Android	B-System
apps	O
,	O
and	O
many	O
of	O
them	O
offer	O
users	O
their	O
own	O
app	O
stores	O
containing	O
programs	O
that	O
can	O
not	O
be	O
found	O
in	O
Google	B-System
Play	I-System
.	O

The	O
purely	O
nominal	O
control	O
over	O
the	O
applications	O
uploaded	O
to	O
these	O
stores	O
means	O
attackers	O
can	O
conceal	O
Trojans	O
in	O
apps	O
made	O
to	O
look	O
like	O
innocent	O
games	O
or	O
utilities	O
.	O

Distribution	O
via	O
botnets	O
.	O

As	O
a	O
rule	O
,	O
bots	O
self-proliferate	O
by	O
sending	O
out	O
text	O
messages	O
with	O
a	O
malicious	O
link	O
to	O
addresses	O
in	O
the	O
victim	O
’	O
s	O
address	O
book	O
.	O

We	O
also	O
registered	O
one	O
episode	O
of	O
mobile	O
malware	O
spreading	O
via	O
a	O
third-party	O
botnet	O
.	O

Resistance	O
to	O
anti-malware	O
protection	O
The	O
ability	O
of	O
malicious	O
software	O
to	O
operate	O
continuously	O
on	O
the	O
victim	O
’	O
s	O
mobile	O
device	O
is	O
an	O
important	O
aspect	O
of	O
its	O
development	O
.	O

The	O
longer	O
a	O
Trojan	O
“	O
lives	O
”	O
on	O
a	O
smartphone	O
,	O
the	O
more	O
money	O
it	O
will	O
make	O
for	O
the	O
owner	O
.	O

This	O
is	O
an	O
area	O
where	O
virus	O
writers	O
are	O
actively	O
working	O
,	O
resulting	O
in	O
a	O
large	O
number	O
of	O
technological	O
innovations	O
.	O

Criminals	O
are	O
increasingly	O
using	O
obfuscation	O
,	O
the	O
deliberate	O
act	O
of	O
creating	O
complex	O
code	O
to	O
make	O
it	O
difficult	O
to	O
analyze	O
.	O

The	O
more	O
complex	O
the	O
obfuscation	O
,	O
the	O
longer	O
it	O
will	O
take	O
an	O
antivirus	O
solution	O
to	O
neutralize	O
the	O
malicious	O
code	O
.	O

Tellingly	O
,	O
current	O
virus	O
writers	O
have	O
mastered	O
commercial	O
obfuscators	O
.	O

This	O
implies	O
they	O
have	O
made	O
considerable	O
investments	O
.	O

For	O
example	O
,	O
one	O
commercial	O
obfuscator	O
,	O
which	O
cost	O
€350	O
,	O
was	O
used	O
for	O
Trojans	O
and	O
Opfak.bo	B-Malware
Obad.a	I-Malware
Android	O
vulnerabilities	O
are	O
used	O
by	O
criminals	O
for	O
three	O
reasons	O
:	O
to	O
bypass	O
the	O
code	O
integrity	O
check	O
when	O
installing	O
an	O
application	O
(	O
vulnerability	O
Master	O
Key	O
)	O
;	O
to	O
enhance	O
the	O
rights	O
of	O
malicious	O
applications	O
,	O
considerably	O
extending	O
their	O
capabilities	O
;	O
and	O
to	O
make	O
it	O
more	O
difficult	O
to	O
remove	O
malware	O
.	O

For	O
example	O
,	O
Svpeng	B-Malware
uses	O
a	O
previously	O
unknown	O
vulnerability	O
to	O
protect	O
itself	O
from	O
being	O
removed	O
manually	O
or	O
by	O
the	O
antivirus	O
program	O
.	O

Cybercriminals	O
also	O
exploit	O
the	O
Master	B-Vulnerability
Key	I-Vulnerability
vulnerability	I-Vulnerability
and	O
have	O
learned	O
to	O
embed	O
unsigned	O
executable	O
files	O
in	O
Android	B-System
installation	O
packages	O
.	O

Digital	O
signature	O
verification	O
can	O
be	O
bypassed	O
by	O
giving	O
the	O
malicious	O
file	O
exactly	O
the	O
same	O
name	O
as	O
a	O
legitimate	O
file	O
and	O
placing	O
it	O
on	O
the	O
same	O
level	O
in	O
the	O
archive	O
.	O

The	O
system	O
verifies	O
the	O
signature	O
of	O
the	O
legitimate	O
file	O
while	O
installing	O
the	O
malicious	O
file	O
.	O

Unfortunately	O
,	O
there	O
is	O
a	O
specific	O
feature	O
of	O
Android	O
vulnerabilities	O
that	O
means	O
it	O
is	O
only	O
possible	O
to	O
get	O
rid	O
of	O
them	O
by	O
receiving	O
an	O
update	O
from	O
the	O
device	O
manufacturers	O
.	O

However	O
,	O
many	O
users	O
are	O
in	O
no	O
hurry	O
to	O
update	O
the	O
operating	O
systems	O
of	O
their	O
products	O
.	O

If	O
a	O
smartphone	O
or	O
tablet	O
was	O
released	O
more	O
than	O
a	O
year	O
ago	O
,	O
it	O
is	O
probably	O
no	O
longer	O
supported	O
by	O
the	O
manufacturer	O
and	O
patching	O
of	O
vulnerabilities	O
is	O
no	O
longer	O
provided	O
.	O

In	O
that	O
case	O
,	O
the	O
only	O
help	O
comes	O
from	O
an	O
antivirus	O
solution	O
,	O
for	O
example	O
,	O
Kaspersky	B-System
Internet	I-System
Security	I-System
for	O
Android	B-System
.	O

Embedding	O
malicious	O
code	O
in	O
legitimate	O
programs	O
helps	O
conceal	O
infections	O
from	O
the	O
victim	O
.	O

Of	O
course	O
,	O
this	O
does	O
not	O
mean	O
the	O
digital	O
signature	O
of	O
the	O
software	O
developer	O
can	O
be	O
used	O
.	O

However	O
,	O
due	O
to	O
the	O
absence	O
of	O
certification	O
centers	O
verifying	O
the	O
digital	O
signatures	O
of	O
Android	O
programs	O
,	O
nothing	O
prevents	O
criminals	O
from	O
adding	O
their	O
own	O
signature	O
.	O

As	O
a	O
result	O
,	O
a	O
copy	O
of	O
Angry	B-System
Birds	I-System
installed	O
from	O
an	O
unofficial	O
app	O
store	O
or	O
downloaded	O
from	O
a	O
forum	O
could	O
easily	O
contain	O
malicious	O
functionality	O
.	O

Capabilities	O
and	O
functionality	O
In	O
2013	O
,	O
we	O
detected	O
several	O
technological	O
innovations	O
developed	O
and	O
used	O
by	O
criminals	O
in	O
their	O
malicious	O
software	O
.	O

Below	O
are	O
descriptions	O
of	O
some	O
of	O
the	O
most	O
interesting	O
.	O

Control	O
of	O
malware	O
from	O
a	O
single	O
center	O
provides	O
maximum	O
flexibility	O
.	O

Botnets	O
can	O
make	O
considerably	O
more	O
money	O
than	O
autonomous	O
Trojans	O
.	O

It	O
comes	O
as	O
no	O
surprise	O
then	O
that	O
many	O
SMS-Trojans	O
include	O
bot	O
functionality	O
.	O

According	O
to	O
our	O
estimates	O
,	O
about	O
60	O
%	O
of	O
mobile	O
malware	O
are	O
elements	O
of	O
both	O
large	O
and	O
small	O
mobile	O
botnets	O
.	O

By	O
using	O
Google	B-System
Cloud	I-System
Messaging	I-System
botnet	O
owners	O
can	O
operate	O
without	O
a	O
C	O
&	O
C	O
server	O
,	O
thus	O
eliminating	O
the	O
threat	O
of	O
the	O
botnet	O
being	O
detected	O
and	O
blocked	O
by	O
law	O
enforcement	O
authorities	O
.	O

Google	B-System
Cloud	I-System
Messaging	I-System
is	O
designed	O
to	O
send	O
short	O
message	O
(	O
up	O
to	O
4	O
KB	O
)	O
to	O
mobile	O
devices	O
via	O
Google	B-Organization
services	O
.	O

The	O
developer	O
simply	O
has	O
to	O
register	O
and	O
receive	O
a	O
unique	O
ID	O
for	O
his	O
applications	O
.	O

The	O
commands	O
received	O
via	O
GCM	B-System
can	O
not	O
be	O
blocked	O
immediately	O
on	O
an	O
infected	O
device	O
.	O

We	O
have	O
detected	O
several	O
malicious	O
programs	O
using	O
GCM	B-System
for	O
command	O
and	O
control	O
–	O
the	O
widespread	O
Trojan-SMS.AndroidOS.FakeInst.a	B-Malware
,	O
Trojan-SMS.AndroidOS.Agent.ao	B-Malware
,	O
and	O
Trojan-SMS.AndroidOS.OpFake.a	B-Malware
among	O
others	O
.	O

Google	B-Organization
is	O
actively	O
combating	O
this	O
use	O
of	O
the	O
service	O
,	O
responding	O
quickly	O
to	O
reports	O
from	O
antivirus	O
companies	O
and	O
blocking	O
the	O
IDs	O
of	O
cybercriminals	O
.	O

Attacks	O
on	O
Windows	B-System
XP	I-System
allows	O
mobile	O
malware	O
to	O
infect	O
a	O
PC	O
after	O
connecting	O
a	O
smartphone	O
or	O
tablet	O
.	O

In	O
early	O
2013	O
we	O
detected	O
two	O
identical	O
applications	O
on	O
Google	B-System
Play	I-System
that	O
were	O
allegedly	O
designed	O
for	O
cleaning	O
the	O
operating	O
system	O
of	O
Android-based	B-System
devices	O
from	O
unnecessary	O
processes	O
.	O

In	O
fact	O
,	O
the	O
applications	O
are	O
designed	O
to	O
download	O
the	O
autorun.inf	B-Indicator
file	I-Indicator
,	O
an	O
icon	O
file	O
and	O
the	O
win32-Trojan	B-System
file	O
,	O
which	O
the	O
mobile	O
malicious	O
program	O
locates	O
in	O
the	O
root	O
directory	O
of	O
an	O
SD	B-System
card	I-System
.	O

On	O
connecting	O
a	O
smartphone	O
in	O
the	O
USB	B-System
drive	I-System
emulation	O
mode	O
to	O
a	O
computer	O
running	O
Windows	B-System
XP	I-System
,	O
the	O
system	O
automatically	O
starts	O
the	O
Trojan	O
(	O
if	O
AutoPlay	O
on	O
the	O
external	O
media	O
is	O
not	O
disabled	O
)	O
and	O
is	O
infected	O
.	O

The	O
Trojan	O
allows	O
the	O
criminals	O
to	O
remotely	O
control	O
the	O
victim	O
’	O
s	O
computer	O
and	O
is	O
capable	O
of	O
recording	O
sound	O
from	O
a	O
microphone	O
.	O

We	O
would	O
like	O
to	O
emphasize	O
that	O
this	O
method	O
of	O
attack	O
only	O
works	O
on	O
Windows	B-System
XP	I-System
and	O
Android	B-System
versions	O
prior	O
to	O
2.2	O
.	O

The	O
most	O
advanced	O
mobile	O
malicious	O
programs	O
today	O
are	O
Trojans	O
targeting	O
users	O
’	O
bank	O
accounts	O
–	O
the	O
most	O
attractive	O
source	O
of	O
criminal	O
earnings	O
.	O

Trend	O
of	O
the	O
year	O
:	O
mobile	O
banking	O
Trojans	O
2013	O
was	O
marked	O
by	O
a	O
rapid	O
rise	O
in	O
the	O
number	O
of	O
Android	B-System
banking	O
Trojans	O
.	O

The	O
cyber	O
industry	O
of	O
mobile	O
malware	O
is	O
becoming	O
more	O
focused	O
on	O
making	O
profits	O
more	O
effectively	O
,	O
i.e.	O
,	O
mobile	O
phishing	O
,	O
theft	O
of	O
credit	O
card	O
information	O
,	O
money	O
transfers	O
from	O
bank	O
cards	O
to	O
mobile	O
phones	O
and	O
from	O
phones	O
to	O
the	O
criminalas	O
’	O
e-wallets	O
.	O

Cybercriminals	O
have	O
become	O
obsessed	O
by	O
this	O
method	O
of	O
illegal	O
earnings	O
:	O
at	O
the	O
beginning	O
of	O
the	O
year	O
we	O
knew	O
only	O
67	O
banking	O
Trojans	O
,	O
but	O
by	O
the	O
end	O
of	O
the	O
year	O
there	O
were	O
already	O
1321	O
unique	O
samples	O
.	O

Kaspersky	B-System
Lab	I-System
mobile	O
products	O
prevented	O
2,500	O
infections	O
by	O
banking	O
Trojans	O
.	O

mobile_treats_2013_04s	O
The	O
number	O
of	O
mobile	O
banking	O
Trojans	O
in	O
our	O
collection	O
Mobile	O
banking	O
Trojans	O
can	O
run	O
together	O
with	O
Win-32	B-System
Trojans	O
to	O
bypass	O
the	O
two-factor	O
authentication	O
–	O
mTAN	O
theft	O
(	O
the	O
theft	O
of	O
banking	O
verification	O
codes	O
that	O
banks	O
send	O
their	O
customers	O
in	O
SMS	O
messages	O
)	O
.	O

However	O
,	O
in	O
2013	O
,	O
autonomous	O
mobile	O
banking	O
Trojans	O
developed	O
further	O
.	O

Currently	O
,	O
such	O
Trojans	O
attack	O
a	O
limited	O
number	O
of	O
bank	O
customers	O
,	O
but	O
it	O
is	O
expected	O
that	O
cybercriminals	O
will	O
invent	O
new	O
techniques	O
that	O
will	O
allow	O
them	O
to	O
expand	O
the	O
number	O
and	O
the	O
geography	O
of	O
potential	O
victims	O
.	O

mobile_treats_2013_05s	O
Infections	O
caused	O
by	O
mobile	O
banking	O
programs	O
Today	O
,	O
the	O
majority	O
of	O
banking	O
Trojan	O
attacks	O
affect	O
users	O
in	O
Russia	O
and	O
the	O
CIS	O
.	O

However	O
,	O
this	O
situation	O
will	O
not	O
last	O
long	O
:	O
given	O
the	O
cybercriminals	O
’	O
interest	O
in	O
user	O
bank	O
accounts	O
,	O
the	O
activity	O
of	O
mobile	O
banking	O
Trojans	O
is	O
expected	O
to	O
grow	O
in	O
other	O
countries	O
in	O
2014	O
.	O

As	O
mentioned	O
above	O
,	O
banking	O
Trojans	O
are	O
perhaps	O
the	O
most	O
complex	O
of	O
all	O
mobile	O
threats	O
,	O
and	O
Svpeng	B-Malware
is	O
one	O
of	O
the	O
most	O
striking	O
examples	O
.	O

Svpeng	B-Malware
In	O
mid-July	O
,	O
we	O
detected	O
Trojan-SMS.AndroidOS.Svpeng.a	B-Malware
which	O
,	O
unlike	O
its	O
SMS	O
Trojan	O
counterparts	O
,	O
is	O
focused	O
on	O
stealing	O
money	O
from	O
the	O
victiim	O
’	O
s	O
bank	O
account	O
rather	O
than	O
from	O
his	O
mobile	O
phone	O
.	O

It	O
can	O
not	O
act	O
independently	O
and	O
operates	O
strictly	O
in	O
accordance	O
with	O
commands	O
received	O
from	O
the	O
C	O
&	O
C	O
server	O
.	O

This	O
malicious	O
program	O
spreads	O
via	O
SMS	O
spam	O
and	O
from	O
compromised	O
legitimate	O
sites	O
that	O
redirect	O
mobile	O
users	O
to	O
a	O
malicious	O
resource	O
.	O

There	O
the	O
user	O
is	O
prompted	O
to	O
download	O
and	O
install	O
a	O
Trojan	O
imitating	O
an	O
Adobe	B-System
Flash	I-System
Player	I-System
update	O
.	O

Svpeng	B-Malware
is	O
capable	O
of	O
doing	O
lots	O
of	O
things	O
.	O

It	O
collects	O
information	O
about	O
the	O
smartphone	O
(	O
IMEI	O
,	O
country	O
,	O
service	O
provider	O
,	O
operating	O
system	O
language	O
)	O
and	O
sends	O
it	O
to	O
the	O
host	O
via	O
the	O
HTTP	O
POST	O
request	O
.	O

This	O
appears	O
to	O
be	O
necessary	O
to	O
determine	O
the	O
number	O
of	O
banks	O
the	O
victim	O
may	O
use	O
.	O

Svpeng	B-Malware
is	O
only	O
currently	O
attacking	O
clients	O
of	O
Russian	O
banks	O
.	O

Typically	O
,	O
however	O
,	O
cybercriminals	O
first	O
test-run	O
a	O
technology	O
on	O
the	O
Russian	O
sector	O
of	O
the	O
Internet	O
and	O
then	O
roll	O
it	O
out	O
globally	O
,	O
attacking	O
users	O
in	O
other	O
countries	O
.	O

It	O
steals	O
SMS	O
messages	O
and	O
information	O
about	O
voice	O
calls	O
.	O

It	O
helps	O
the	O
attacker	O
find	O
out	O
which	O
banks	O
the	O
owner	O
of	O
the	O
smartphone	O
calls	O
–	O
the	O
Trojan	O
receives	O
a	O
list	O
of	O
bank	O
phone	O
numbers	O
from	O
its	O
C	O
&	O
C	O
server	O
.	O

It	O
steals	O
money	O
from	O
the	O
victim	O
’	O
s	O
bank	O
account	O
.	O

In	O
Russia	O
,	O
some	O
major	O
banks	O
offer	O
their	O
clients	O
a	O
special	O
service	O
that	O
allows	O
them	O
to	O
transfer	O
money	O
from	O
their	O
bank	O
card	O
to	O
their	O
mobile	O
phone	O
account	O
.	O

Customers	O
have	O
to	O
send	O
a	O
set	O
text	O
message	O
from	O
their	O
phone	O
to	O
a	O
specific	O
bank	O
number	O
.	O

Svpeng	B-Malware
sends	O
the	O
corresponding	O
messages	O
to	O
the	O
SMS	O
services	O
of	O
two	O
banks	O
.	O

Svpeng	B-Malware
does	O
this	O
to	O
check	O
if	O
the	O
cards	O
from	O
these	O
banks	O
are	O
attached	O
to	O
the	O
number	O
of	O
the	O
infected	O
phone	O
and	O
to	O
find	O
out	O
the	O
account	O
balance	O
.	O

If	O
the	O
phone	O
is	O
attached	O
to	O
a	O
bank	O
card	O
,	O
commands	O
are	O
sent	O
from	O
the	O
C	O
&	O
C	O
server	O
with	O
instructions	O
to	O
transfer	O
money	O
from	O
the	O
user	O
’	O
s	O
bank	O
account	O
to	O
his/her	O
mobile	O
account	O
.	O

The	O
cybercriminals	O
then	O
send	O
this	O
money	O
to	O
a	O
digital	O
wallet	O
or	O
to	O
a	O
premium	O
number	O
and	O
cash	O
it	O
in	O
.	O

It	O
steals	O
logins	O
and	O
passwords	O
to	O
online	O
banking	O
accounts	O
by	O
substituting	O
he	O
window	O
displayed	O
by	O
the	O
bank	O
application	O
.	O

Currently	O
,	O
this	O
only	O
affects	O
Russian	O
banks	O
,	O
but	O
the	O
technology	O
behind	O
Svpeng	B-Malware
could	O
easily	O
be	O
used	O
to	O
target	O
other	O
banking	O
applications	O
.	O

It	O
steals	O
bank	O
card	O
information	O
(	O
the	O
number	O
,	O
the	O
expiry	O
date	O
,	O
CVC2/CVV2	O
)	O
imitating	O
the	O
process	O
of	O
registering	O
the	O
bank	O
card	O
with	O
Google	B-System
Play	I-System
.	O

If	O
the	O
user	O
has	O
launched	O
Play	B-System
Market	I-System
,	O
the	O
Trojan	O
intercepts	O
the	O
event	O
and	O
displays	O
a	O
window	O
on	O
top	O
of	O
the	O
Google	B-System
Play	I-System
window	O
,	O
prompting	O
the	O
user	O
to	O
enter	O
his/her	O
bank	O
card	O
details	O
in	O
the	O
fake	O
window	O
.	O

The	O
data	O
entered	O
by	O
the	O
user	O
is	O
sent	O
to	O
the	O
cybercriminals	O
.	O

mobile_treats_2013_06s	O
It	O
extorts	O
money	O
from	O
users	O
by	O
threatening	O
to	O
block	O
the	O
smartphone	O
:	O
it	O
displays	O
a	O
message	O
demanding	O
$	O
500	O
to	O
unblock	O
the	O
device	O
.	O

In	O
actual	O
fact	O
,	O
the	O
Trojan	O
does	O
not	O
block	O
anything	O
and	O
the	O
phone	O
can	O
be	O
used	O
without	O
any	O
problems	O
.	O

It	O
hides	O
traces	O
of	O
its	O
activity	O
by	O
masking	O
the	O
outgoing	O
and	O
incoming	O
text	O
messages	O
and	O
blocking	O
calls	O
and	O
messages	O
from	O
numbers	O
belonging	O
to	O
the	O
bank	O
.	O

The	O
Trojan	O
gets	O
the	O
list	O
of	O
bank	O
phone	O
numbers	O
from	O
its	O
C	O
&	O
C	O
server	O
.	O

It	O
protects	O
itself	O
from	O
deletion	O
by	O
requesting	O
Device	O
Administrator	O
rights	O
during	O
the	O
installation	O
.	O

As	O
a	O
result	O
,	O
the	O
Trojan	O
delete	O
button	O
in	O
the	O
list	O
of	O
applications	O
becomes	O
inactive	O
,	O
which	O
may	O
cause	O
problems	O
for	O
inexperienced	O
users	O
.	O

It	O
is	O
impossible	O
to	O
deprive	O
it	O
of	O
these	O
rights	O
without	O
the	O
use	O
of	O
specialized	O
tools	O
(	O
such	O
as	O
Kaspersky	B-System
Internet	I-System
Security	I-System
for	O
Android	B-System
)	O
.	O

To	O
protect	O
itself	O
from	O
being	O
removed	O
,	O
Svpeng	B-Malware
uses	O
a	O
previously	O
unknown	O
vulnerability	O
in	O
Android	B-System
.	O

It	O
uses	O
the	O
same	O
trick	O
to	O
prevent	O
the	O
smartphone	O
from	O
being	O
returned	O
to	O
its	O
factory	O
settings	O
.	O

The	O
Trojan	O
is	O
distributed	O
in	O
Russia	O
and	O
CIS	O
countries	O
.	O

But	O
,	O
as	O
we	O
have	O
already	O
mentioned	O
,	O
the	O
criminals	O
could	O
easily	O
turn	O
their	O
attention	O
to	O
users	O
in	O
other	O
countries	O
.	O

Perkele	B-Malware
and	O
Wroba	B-Malware
Foreign	O
users	O
have	O
also	O
been	O
on	O
the	O
receiving	O
end	O
of	O
several	O
malicious	O
innovations	O
targeting	O
bank	O
accounts	O
.	O

The	O
Perkele	B-Malware
Android	O
Trojan	O
not	O
only	O
attacks	O
Russian	O
users	O
but	O
also	O
clients	O
of	O
several	O
European	O
banks	O
.	O

It	O
is	O
of	O
interest	O
primarily	O
because	O
it	O
operates	O
in	O
conjunction	O
with	O
various	O
banking	O
win32-Trojans	B-System
.	O

Its	O
main	O
task	O
is	O
to	O
bypass	O
the	O
two-factor	O
authentication	O
of	O
the	O
client	O
in	O
the	O
online	O
banking	O
system	O
.	O

Due	O
to	O
the	O
specific	O
nature	O
of	O
its	O
activity	O
,	O
Perkele	B-Malware
is	O
distributed	O
in	O
a	O
rather	O
unusual	O
way	O
.	O

When	O
a	O
user	O
enters	O
an	O
Internet	O
banking	O
site	O
on	O
a	O
computer	O
infected	O
by	O
banking	O
malware	O
(	O
ZeuS	B-Malware
,	O
Citadel	B-Malware
)	O
,	O
a	O
request	O
about	O
the	O
smartphone	O
number	O
and	O
type	O
of	O
operating	O
system	O
is	O
injected	O
into	O
the	O
code	O
of	O
the	O
authentication	O
page	O
.	O

This	O
data	O
is	O
immediately	O
sent	O
to	O
the	O
cybercriminals	O
and	O
the	O
computer	O
displays	O
the	O
QR	O
code	O
containing	O
a	O
link	O
to	O
the	O
alleged	O
certificate	O
of	O
the	O
online	O
banking	O
system	O
.	O

After	O
scanning	O
the	O
QR	O
code	O
and	O
installing	O
a	O
component	O
downloaded	O
from	O
the	O
link	O
,	O
the	O
user	O
infects	O
his	O
smartphone	O
with	O
the	O
Trojan	O
program	O
that	O
boasts	O
functionality	O
that	O
is	O
of	O
great	O
interest	O
to	O
the	O
attackers	O
.	O

Perkele	B-Malware
intercepts	O
mTANs	O
(	O
confirmation	O
codes	O
for	O
banking	O
operations	O
)	O
sent	O
by	O
the	O
bank	O
via	O
text	O
message	O
.	O

By	O
using	O
the	O
login	O
and	O
password	O
stolen	O
from	O
the	O
browser	O
,	O
the	O
Windows	O
Trojan	O
initiates	O
a	O
fake	O
transaction	O
while	O
Perkele	B-Malware
intercepts	O
(	O
via	O
the	O
C	O
&	O
C	O
server	O
)	O
the	O
mTAN	O
sent	O
by	O
the	O
bank	O
to	O
the	O
user	O
.	O

Money	O
then	O
disappears	O
from	O
the	O
victim	O
’	O
s	O
account	O
and	O
is	O
cashed	O
in	O
without	O
the	O
owner	O
’	O
s	O
knowledge	O
.	O

The	O
Korean	O
malware	O
Wroba	B-Malware
,	O
in	O
addition	O
to	O
the	O
traditional	O
vector	O
of	O
infection	O
via	O
file-sharing	O
services	O
,	O
spreads	O
via	O
alternative	O
app	O
stores	O
.	O

Once	O
it	O
infects	O
a	O
device	O
,	O
Wroba	B-Malware
behaves	O
very	O
aggressively	O
.	O

It	O
searches	O
for	O
mobile	O
banking	O
applications	O
,	O
removes	O
them	O
and	O
uploads	O
counterfeit	O
versions	O
.	O

From	O
the	O
outside	O
,	O
they	O
are	O
indistinguishable	O
from	O
the	O
legitimate	O
applications	O
.	O

However	O
,	O
they	O
possess	O
no	O
banking	O
functions	O
,	O
and	O
merely	O
steal	O
the	O
logins	O
and	O
passwords	O
entered	O
by	O
users	O
.	O

ViperRAT	B-Malware
:	O
The	O
Mobile	O
APT	O
Targeting	O
The	O
Israeli	B-Organization
Defense	I-Organization
Force	I-Organization
That	O
Should	O
Be	O
On	O
Your	O
Radar	O
February	O
16	O
,	O
2017	O
ViperRAT	B-Malware
is	O
an	O
active	O
,	O
advanced	O
persistent	O
threat	O
(	O
APT	O
)	O
that	O
sophisticated	O
threat	O
actors	O
are	O
actively	O
using	O
to	O
target	O
and	O
spy	O
on	O
the	O
Israeli	B-Organization
Defense	I-Organization
Force.The	I-Organization
threat	O
actors	O
behind	O
the	O
ViperRAT	B-Malware
surveillanceware	O
collect	O
a	O
significant	O
amount	O
of	O
sensitive	O
information	O
off	O
of	O
the	O
device	O
,	O
and	O
seem	O
most	O
interested	O
in	O
exfiltrating	O
images	O
and	O
audio	O
content	O
.	O

The	O
attackers	O
are	O
also	O
hijacking	O
the	O
device	O
camera	O
to	O
take	O
pictures	O
.	O

Using	O
data	O
collected	O
from	O
the	O
Lookout	B-Organization
global	O
sensor	O
network	O
,	O
the	O
Lookout	O
research	O
team	O
was	O
able	O
to	O
gain	O
unique	O
visibility	O
into	O
the	O
ViperRAT	B-Malware
malware	O
,	O
including	O
11	O
new	O
,	O
unreported	O
applications	O
.	O

We	O
also	O
discovered	O
and	O
analyzed	O
live	O
,	O
misconfigured	O
malicious	O
command	O
and	O
control	O
servers	O
(	O
C2	O
)	O
,	O
from	O
which	O
we	O
were	O
able	O
to	O
identify	O
how	O
the	O
attacker	O
gets	O
new	O
,	O
infected	O
apps	O
to	O
secretly	O
install	O
and	O
the	O
types	O
of	O
activities	O
they	O
are	O
monitoring	O
.	O

In	O
addition	O
,	O
we	O
uncovered	O
the	O
IMEIs	O
of	O
the	O
targeted	O
individuals	O
(	O
IMEIs	O
will	O
not	O
be	O
shared	O
publicly	O
for	O
the	O
privacy	O
and	O
safety	O
of	O
the	O
victims	O
)	O
as	O
well	O
as	O
the	O
types	O
of	O
exfiltrated	O
content	O
.	O

In	O
aggregate	O
,	O
the	O
type	O
of	O
information	O
stolen	O
could	O
let	O
an	O
attacker	O
know	O
where	O
a	O
person	O
is	O
,	O
with	O
whom	O
they	O
are	O
associated	O
(	O
including	O
contacts	O
’	O
profile	O
photos	O
)	O
,	O
the	O
messages	O
they	O
are	O
sending	O
,	O
the	O
websites	O
they	O
visit	O
and	O
search	O
history	O
,	O
screenshots	O
that	O
reveal	O
data	O
from	O
other	O
apps	O
on	O
the	O
device	O
,	O
the	O
conversations	O
they	O
have	O
in	O
the	O
presence	O
of	O
the	O
device	O
,	O
and	O
a	O
myriad	O
of	O
images	O
including	O
anything	O
at	O
which	O
device	O
’	O
s	O
camera	O
is	O
pointed	O
.	O

Lookout	B-Organization
has	O
determined	O
ViperRAT	B-Malware
is	O
a	O
very	O
sophisticated	O
threat	O
that	O
adds	O
to	O
the	O
mounting	O
evidence	O
that	O
targeted	O
mobile	O
attacks	O
against	O
governments	O
and	O
business	O
is	O
a	O
real	O
problem	O
.	O

Lookout	B-Organization
researchers	O
have	O
been	O
tracking	O
this	O
threat	O
for	O
the	O
last	O
month	O
.	O

Given	O
that	O
this	O
is	O
an	O
active	O
threat	O
,	O
we	O
’	O
ve	O
been	O
working	O
behind-the-scenes	O
with	O
our	O
customers	O
to	O
ensure	O
both	O
personal	O
and	O
enterprise	O
customers	O
are	O
protected	O
from	O
this	O
threat	O
and	O
only	O
decided	O
to	O
come	O
forward	O
with	O
this	O
information	O
after	O
the	O
research	O
team	O
at	O
Kaspersky	B-Organization
released	O
a	O
report	O
earlier	O
today	O
.	O

Additionally	O
,	O
we	O
have	O
determined	O
that	O
though	O
original	O
reports	O
of	O
this	O
story	O
attribute	O
this	O
surveillanceware	O
tool	O
to	O
Hamas	B-Organization
,	O
this	O
may	O
not	O
be	O
the	O
case	O
,	O
as	O
we	O
demonstrate	O
below	O
.	O

The	O
increasing	O
sophistication	O
of	O
surveillanceware	O
The	O
structure	O
of	O
the	O
surveillanceware	O
indicates	O
it	O
is	O
very	O
sophisticated	O
.	O

Analysis	O
indicates	O
there	O
are	O
currently	O
two	O
distinct	O
variants	O
of	O
ViperRAT	B-Malware
.	O

The	O
first	O
variant	O
is	O
a	O
“	O
first	O
stage	O
application	O
,	O
”	O
that	O
performs	O
basic	O
profiling	O
of	O
a	O
device	O
,	O
and	O
under	O
certain	O
conditions	O
attempts	O
to	O
download	O
and	O
install	O
a	O
much	O
more	O
comprehensive	O
surveillanceware	O
component	O
,	O
which	O
is	O
the	O
second	O
variant	O
.	O

The	O
first	O
variant	O
involves	O
social	O
engineering	O
the	O
target	O
into	O
downloading	O
a	O
trojanized	O
app	O
.	O

Previous	O
reports	O
alleged	O
this	O
surveillanceware	O
tool	O
was	O
deployed	O
using	O
‘	O
honey	O
traps	O
’	O
where	O
the	O
actor	O
behind	O
it	O
would	O
reach	O
out	O
to	O
targets	O
via	O
fake	O
social	O
media	O
profiles	O
of	O
young	O
women	O
.	O

After	O
building	O
an	O
initial	O
rapport	O
with	O
targets	O
,	O
the	O
actors	O
behind	O
these	O
social	O
media	O
accounts	O
would	O
instruct	O
victims	O
to	O
install	O
an	O
additional	O
app	O
for	O
easier	O
communication	O
.	O

Specifically	O
,	O
Lookout	B-Organization
determined	O
these	O
were	O
trojanized	O
versions	O
of	O
the	O
apps	O
SR	B-System
Chat	I-System
and	O
YeeCall	B-System
Pro	I-System
.	O

We	O
also	O
uncovered	O
ViperRAT	B-Malware
in	O
a	O
billiards	O
game	O
,	O
an	O
Israeli	O
Love	O
Songs	O
player	O
,	O
and	O
a	O
Move	O
To	O
iOS	B-System
app	O
.	O

The	O
second	O
stage	O
The	O
second	O
stage	O
apps	O
contain	O
the	O
surveillanceware	O
capabilities	O
.	O

Lookout	B-Organization
uncovered	O
nine	O
secondary	O
payload	O
applications	O
:	O
*	O
These	O
apps	O
have	O
not	O
been	O
previously	O
reported	O
and	O
were	O
discovered	O
using	O
data	O
from	O
the	O
Lookout	B-Organization
global	O
sensor	O
network	O
,	O
which	O
collects	O
app	O
and	O
device	O
information	O
from	O
over	O
100	O
million	O
sensors	O
to	O
provide	O
researchers	O
and	O
customers	O
with	O
a	O
holistic	O
look	O
at	O
the	O
mobile	O
threat	O
ecosystem	O
today	O
.	O

Naming	O
additional	O
payload	O
applications	O
as	O
system	O
updates	O
is	O
a	O
clever	O
technique	O
used	O
by	O
malware	O
authors	O
to	O
trick	O
victims	O
into	O
believing	O
a	O
threat	O
isn	O
’	O
t	O
present	O
on	O
their	O
device	O
.	O

ViperRAT	B-Malware
takes	O
this	O
one	O
step	O
further	O
by	O
using	O
its	O
dropper	O
app	O
to	O
identify	O
an	O
appropriate	O
second	O
stage	O
‘	O
update	O
’	O
that	O
may	O
go	O
unnoticed	O
.	O

For	O
example	O
,	O
if	O
a	O
victim	O
has	O
Viber	B-System
on	O
their	O
device	O
,	O
it	O
will	O
choose	O
to	O
retrieve	O
the	O
Viber	B-System
Update	I-System
second	O
stage	O
.	O

If	O
he	O
doesn	O
’	O
t	O
have	O
Viber	O
,	O
the	O
generically-named	O
System	O
Updates	O
app	O
gets	O
downloaded	O
and	O
installed	O
instead	O
.	O

What	O
was	O
taken	O
The	O
actors	O
behind	O
ViperRAT	B-Malware
seem	O
to	O
be	O
particularly	O
interested	O
in	O
image	O
data	O
.	O

We	O
were	O
able	O
to	O
identify	O
that	O
8,929	O
files	O
had	O
been	O
exfiltrated	O
from	O
compromised	O
devices	O
and	O
that	O
the	O
overwhelming	O
majority	O
of	O
these	O
,	O
97	O
percent	O
,	O
were	O
highly	O
likely	O
encrypted	O
images	O
taken	O
using	O
the	O
device	O
camera	O
.	O

We	O
also	O
observed	O
automatically	O
generated	O
files	O
on	O
the	O
C2	O
,	O
indicating	O
the	O
actor	O
behind	O
this	O
campaign	O
also	O
issues	O
commands	O
to	O
search	O
for	O
and	O
exfiltrate	O
PDF	O
and	O
Office	O
documents	O
.	O

This	O
should	O
be	O
highly	O
alarming	O
to	O
any	O
government	O
agency	O
or	O
enterprise	O
.	O

We	O
observed	O
legitimate	O
exfiltrated	O
files	O
of	O
the	O
following	O
types	O
of	O
data	O
:	O
Contact	O
information	O
Compressed	O
recorded	O
audio	O
in	O
the	O
Adaptive	O
Multi-Rate	O
(	O
amr	O
)	O
file	O
format	O
Images	O
captured	O
from	O
the	O
device	O
camera	O
Images	O
stored	O
on	O
both	O
internal	O
device	O
and	O
SDCard	O
storage	O
that	O
are	O
listed	O
in	O
the	O
MediaStore	O
Device	O
geolocation	O
information	O
SMS	O
content	O
Chrome	O
browser	O
search	O
history	O
and	O
bookmarks	O
Call	O
log	O
information	O
Cell	O
tower	O
information	O
Device	O
network	O
metadata	O
;	O
such	O
as	O
phone	O
number	O
,	O
device	O
software	O
version	O
,	O
network	O
country	O
,	O
network	O
operator	O
,	O
SIM	O
country	O
,	O
SIM	O
operator	O
,	O
SIM	O
serial	O
,	O
IMSI	O
,	O
voice	O
mail	O
number	O
,	O
phone	O

type	O
,	O
network	O
type	O
,	O
data	O
state	O
,	O
data	O
activity	O
,	O
call	O
state	O
,	O
SIM	O
state	O
,	O
whether	O
device	O
is	O
roaming	O
,	O
and	O
if	O
SMS	O
is	O
supported	O
.	O

Standard	O
browser	O
search	O
history	O
Standard	O
browser	O
bookmarks	O
Device	O
handset	O
metadata	O
;	O
such	O
as	O
brand	O
,	O
display	O
,	O
hardware	O
,	O
manufacturer	O
,	O
product	O
,	O
serial	O
,	O
radio	O
version	O
,	O
and	O
SDK	O
.	O

Command	O
and	O
control	O
API	O
calls	O
ViperRAT	B-Malware
samples	O
are	O
capable	O
of	O
communicating	O
to	O
C2	O
servers	O
through	O
an	O
exposed	O
API	O
as	O
well	O
as	O
websockets	O
.	O

Below	O
is	O
a	O
collection	O
of	O
API	O
methods	O
and	O
a	O
brief	O
description	O
around	O
their	O
purpose	O
.	O

On	O
attribution	O
Media	O
reporting	O
on	O
ViperRAT	B-Malware
thus	O
far	O
attributes	O
this	O
surveillanceware	O
tool	O
to	O
Hamas	B-Organization
.	O

Israeli	O
media	O
published	O
the	O
first	O
reports	O
about	O
the	O
social	O
networking	O
and	O
social	O
engineering	O
aspects	O
of	O
this	O
campaign	O
.	O

However	O
it	O
’	O
s	O
unclear	O
whether	O
organizations	O
that	O
later	O
reported	O
on	O
ViperRAT	B-Malware
performed	O
their	O
own	O
independent	O
research	O
or	O
simply	O
based	O
their	O
content	O
on	O
the	O
original	O
Israeli	O
report	O
.	O

Hamas	B-Organization
is	O
not	O
widely	O
known	O
for	O
having	O
a	O
sophisticated	O
mobile	O
capability	O
,	O
which	O
makes	O
it	O
unlikely	O
they	O
are	O
directly	O
responsible	O
for	O
ViperRAT	B-Malware
.	O

ViperRAT	B-Malware
has	O
been	O
operational	O
for	O
quite	O
some	O
time	O
,	O
with	O
what	O
appears	O
to	O
be	O
a	O
test	O
application	O
that	O
surfaced	O
in	O
late	O
2015	O
.	O

Many	O
of	O
the	O
default	O
strings	O
in	O
this	O
application	O
are	O
in	O
Arabic	O
,	O
including	O
the	O
name	O
.	O

It	O
is	O
unclear	O
whether	O
this	O
means	O
early	O
samples	O
were	O
targeting	O
Arabic	O
speakers	O
or	O
if	O
the	O
developers	O
behind	O
it	O
are	O
fluent	O
in	O
Arabic	O
.	O

This	O
leads	O
us	O
to	O
believe	O
this	O
is	O
another	O
actor	O
.	O

What	O
this	O
means	O
for	O
you	O
All	O
Lookout	B-Organization
customers	O
are	O
protected	O
from	O
this	O
threat	O
.	O

However	O
,	O
the	O
existence	O
of	O
threats	O
like	O
ViperRAT	B-Malware
and	O
Pegasus	B-Malware
,	O
the	O
most	O
sophisticated	O
piece	O
of	O
mobile	O
surveillanceware	O
we	O
’	O
ve	O
seen	O
to	O
date	O
,	O
are	O
evidence	O
that	O
attackers	O
are	O
targeting	O
mobile	O
devices	O
.	O

Mobile	O
devices	O
are	O
at	O
the	O
frontier	O
of	O
cyber	O
espionage	O
,	O
and	O
other	O
criminal	O
motives	O
.	O

Enterprise	O
and	O
government	O
employees	O
all	O
use	O
these	O
devices	O
in	O
their	O
day-to-day	O
work	O
,	O
which	O
means	O
IT	O
and	O
security	O
leaders	O
within	O
these	O
organizations	O
must	O
prioritize	O
mobile	O
in	O
their	O
security	O
strategies	O
.	O

Check	B-Organization
Point	I-Organization
researchers	O
discovered	O
another	O
widespread	O
malware	O
campaign	O
on	O
Google	B-System
Play	I-System
,	O
Google	B-Organization
’	O
s	O
official	O
app	O
store	O
.	O

The	O
malware	O
,	O
dubbed	O
“	O
Judy	B-Malware
”	O
,	O
is	O
an	O
auto-clicking	O
adware	O
which	O
was	O
found	O
on	O
41	O
apps	O
developed	O
by	O
a	O
Korean	O
company	O
.	O

The	O
malware	O
uses	O
infected	O
devices	O
to	O
generate	O
large	O
amounts	O
of	O
fraudulent	O
clicks	O
on	O
advertisements	O
,	O
generating	O
revenues	O
for	O
the	O
perpetrators	O
behind	O
it	O
.	O

The	O
malicious	O
apps	O
reached	O
an	O
astonishing	O
spread	O
between	O
4.5	O
million	O
and	O
18.5	O
million	O
downloads	O
.	O

Some	O
of	O
the	O
apps	O
we	O
discovered	O
resided	O
on	O
Google	B-System
Play	I-System
for	O
several	O
years	O
,	O
but	O
all	O
were	O
recently	O
updated	O
.	O

It	O
is	O
unclear	O
how	O
long	O
the	O
malicious	O
code	O
existed	O
inside	O
the	O
apps	O
,	O
hence	O
the	O
actual	O
spread	O
of	O
the	O
malware	O
remains	O
unknown	O
.	O

We	O
also	O
found	O
several	O
apps	O
containing	O
the	O
malware	O
,	O
which	O
were	O
developed	O
by	O
other	O
developers	O
on	O
Google	B-System
Play	I-System
.	O

The	O
connection	O
between	O
the	O
two	O
campaigns	O
remains	O
unclear	O
,	O
and	O
it	O
is	O
possible	O
that	O
one	O
borrowed	O
code	O
from	O
the	O
other	O
,	O
knowingly	O
or	O
unknowingly	O
.	O

The	O
oldest	O
app	O
of	O
the	O
second	O
campaign	O
was	O
last	O
updated	O
in	O
April	O
2016	O
,	O
meaning	O
that	O
the	O
malicious	O
code	O
hid	O
for	O
a	O
long	O
time	O
on	O
the	O
Play	B-System
store	I-System
undetected	O
.	O

These	O
apps	O
also	O
had	O
a	O
large	O
amount	O
of	O
downloads	O
between	O
4	O
and	O
18	O
million	O
,	O
meaning	O
the	O
total	O
spread	O
of	O
the	O
malware	O
may	O
have	O
reached	O
between	O
8.5	O
and	O
36.5	O
million	O
users	O
.	O

Similar	O
to	O
previous	O
malware	O
which	O
infiltrated	O
Google	B-System
Play	I-System
,	O
such	O
as	O
FalseGuide	B-Malware
and	O
Skinner	B-Malware
,	O
Judy	O
relies	O
on	O
the	O
communication	O
with	O
its	O
Command	O
and	O
Control	O
server	O
(	O
C	O
&	O
C	O
)	O
for	O
its	O
operation	O
.	O

After	O
Check	B-Organization
Point	I-Organization
notified	O
Google	B-Organization
about	O
this	O
threat	O
,	O
the	O
apps	O
were	O
swiftly	O
removed	O
from	O
the	O
Play	B-System
store	I-System
.	O

How	O
Judy	B-Malware
operates	O
:	O
To	O
bypass	O
Bouncer	B-System
,	O
Google	B-System
Play	I-System
’	O
s	O
protection	O
,	O
the	O
hackers	O
create	O
a	O
seemingly	O
benign	O
bridgehead	O
app	O
,	O
meant	O
to	O
establish	O
connection	O
to	O
the	O
victim	O
’	O
s	O
device	O
,	O
and	O
insert	O
it	O
into	O
the	O
app	O
store	O
.	O

Once	O
a	O
user	O
downloads	O
a	O
malicious	O
app	O
,	O
it	O
silently	O
registers	O
receivers	O
which	O
establish	O
a	O
connection	O
with	O
the	O
C	O
&	O
C	O
server	O
.	O

The	O
server	O
replies	O
with	O
the	O
actual	O
malicious	O
payload	O
,	O
which	O
includes	O
JavaScript	O
code	O
,	O
a	O
user-agent	O
string	O
and	O
URLs	O
controlled	O
by	O
the	O
malware	O
author	O
.	O

The	O
malware	O
opens	O
the	O
URLs	O
using	O
the	O
user	O
agent	O
that	O
imitates	O
a	O
PC	O
browser	O
in	O
a	O
hidden	O
webpage	O
and	O
receives	O
a	O
redirection	O
to	O
another	O
website	O
.	O

Once	O
the	O
targeted	O
website	O
is	O
launched	O
,	O
the	O
malware	O
uses	O
the	O
JavaScript	O
code	O
to	O
locate	O
and	O
click	O
on	O
banners	O
from	O
the	O
Google	B-System
ads	I-System
infrastructure	O
.	O

Upon	O
clicking	O
the	O
ads	O
,	O
the	O
malware	O
author	O
receives	O
payment	O
from	O
the	O
website	O
developer	O
,	O
which	O
pays	O
for	O
the	O
illegitimate	O
clicks	O
and	O
traffic	O
.	O

The	O
JavaScript	O
code	O
locates	O
the	O
targeted	O
ads	O
by	O
searching	O
for	O
iframes	O
which	O
contain	O
ads	O
from	O
Google	B-System
ads	I-System
infrastructure	O
,	O
as	O
shown	O
in	O
the	O
image	O
below	O
:	O
The	O
fraudulent	O
clicks	O
generate	O
a	O
large	O
revenue	O
for	O
the	O
perpetrators	O
,	O
especially	O
since	O
the	O
malware	O
reached	O
a	O
presumably	O
wide	O
spread	O
.	O

Who	O
is	O
behind	O
Judy	B-Malware
?	O

The	O
malicious	O
apps	O
are	O
all	O
developed	O
by	O
a	O
Korean	O
company	O
named	O
Kiniwini	B-Organization
,	O
registered	O
on	O
Google	B-System
Play	I-System
as	O
ENISTUDIO	B-Organization
corp	I-Organization
.	O

The	O
company	O
develops	O
mobile	O
apps	O
for	O
both	O
Android	B-System
and	O
iOS	B-System
platforms	O
.	O

It	O
is	O
quite	O
unusual	O
to	O
find	O
an	O
actual	O
organization	O
behind	O
mobile	O
malware	O
,	O
as	O
most	O
of	O
them	O
are	O
developed	O
by	O
purely	O
malicious	O
actors	O
.	O

It	O
is	O
important	O
to	O
note	O
that	O
the	O
activity	O
conducted	O
by	O
the	O
malware	O
is	O
not	O
borderline	O
advertising	O
,	O
but	O
definitely	O
an	O
illegitimate	O
use	O
of	O
the	O
users	O
’	O
mobile	O
devices	O
for	O
generating	O
fraudulent	O
clicks	O
,	O
benefiting	O
the	O
attackers	O
.	O

In	O
addition	O
to	O
the	O
clicking	O
activity	O
,	O
Judy	B-Malware
displays	O
a	O
large	O
amount	O
of	O
advertisements	O
,	O
which	O
in	O
some	O
cases	O
leave	O
users	O
with	O
no	O
option	O
but	O
clicking	O
on	O
the	O
ad	O
itself	O
.	O

Although	O
most	O
apps	O
have	O
positive	O
ratings	O
,	O
some	O
of	O
the	O
users	O
have	O
noticed	O
and	O
reported	O
Judy	B-Malware
’	O
s	O
suspicious	O
activities	O
,	O
as	O
seen	O
in	O
the	O
images	O
below	O
:	O
As	O
seen	O
in	O
previous	O
malware	O
,	O
such	O
as	O
DressCode	B-Malware
,	O
a	O
high	O
reputation	O
does	O
not	O
necessarily	O
indicate	O
that	O
the	O
app	O
is	O
safe	O
for	O
use	O
.	O

Hackers	O
can	O
hide	O
their	O
apps	O
’	O
real	O
intentions	O
or	O
even	O
manipulate	O
users	O
into	O
leaving	O
positive	O
ratings	O
,	O
in	O
some	O
cases	O
unknowingly	O
.	O

Users	O
can	O
not	O
rely	O
on	O
the	O
official	O
app	O
stores	O
for	O
their	O
safety	O
,	O
and	O
should	O
implement	O
advanced	O
security	O
protections	O
capable	O
of	O
detecting	O
and	O
blocking	O
zero-day	O
mobile	O
malware	O
.	O

PHA	O
Family	O
Highlights	O
:	O
Bread	B-Malware
(	O
and	O
Friends	O
)	O
January	O
9	O
,	O
2020	O
In	O
this	O
edition	O
of	O
our	O
PHA	O
Family	O
Highlights	O
series	O
we	O
introduce	O
Bread	B-Malware
,	O
a	O
large-scale	O
billing	O
fraud	O
family	O
.	O

We	O
first	O
started	O
tracking	O
Bread	B-Malware
(	O
also	O
known	O
as	O
Joker	B-Malware
)	O
in	O
early	O
2017	O
,	O
identifying	O
apps	O
designed	O
solely	O
for	O
SMS	O
fraud	O
.	O

As	O
the	O
Play	B-System
Store	I-System
has	O
introduced	O
new	O
policies	O
and	O
Google	B-System
Play	I-System
Protect	I-System
has	O
scaled	O
defenses	O
,	O
Bread	B-Malware
apps	O
were	O
forced	O
to	O
continually	O
iterate	O
to	O
search	O
for	O
gaps	O
.	O

They	O
have	O
at	O
some	O
point	O
used	O
just	O
about	O
every	O
cloaking	O
and	O
obfuscation	O
technique	O
under	O
the	O
sun	O
in	O
an	O
attempt	O
to	O
go	O
undetected	O
.	O

Many	O
of	O
these	O
samples	O
appear	O
to	O
be	O
designed	O
specifically	O
to	O
attempt	O
to	O
slip	O
into	O
the	O
Play	B-System
Store	I-System
undetected	O
and	O
are	O
not	O
seen	O
elsewhere	O
.	O

In	O
this	O
post	O
,	O
we	O
show	O
how	O
Google	B-System
Play	I-System
Protect	I-System
has	O
defended	O
against	O
a	O
well	O
organized	O
,	O
persistent	O
attacker	O
and	O
share	O
examples	O
of	O
their	O
techniques	O
.	O

TL	O
;	O
DR	O
Google	B-System
Play	I-System
Protect	I-System
detected	O
and	O
removed	O
1.7k	O
unique	O
Bread	B-Malware
apps	O
from	O
the	O
Play	B-System
Store	I-System
before	O
ever	O
being	O
downloaded	O
by	O
users	O
Bread	B-Malware
apps	O
originally	O
performed	O
SMS	O
fraud	O
,	O
but	O
have	O
largely	O
abandoned	O
this	O
for	O
WAP	O
billing	O
following	O
the	O
introduction	O
of	O
new	O
Play	B-System
policies	O
restricting	O
use	O
of	O
the	O
SEND_SMS	O
permission	O
and	O
increased	O
coverage	O
by	O
Google	B-System
Play	I-System
Protect	I-System
More	O
information	O
on	O
stats	O
and	O
relative	O
impact	O
is	O
available	O
in	O
the	O
Android	B-System
Security	O
2018	O
Year	O
in	O
Review	O
report	O
BILLING	O
FRAUD	O
Bread	B-Malware
apps	O
typically	O
fall	O
into	O
two	O
categories	O
:	O
SMS	O
fraud	O
(	O
older	O
versions	O
)	O
and	O
toll	O
fraud	O
(	O
newer	O
versions	O
)	O
.	O

Both	O
of	O
these	O
types	O
of	O
fraud	O
take	O
advantage	O
of	O
mobile	O
billing	O
techniques	O
involving	O
the	O
user	O
’	O
s	O
carrier	O
.	O

SMS	O
Billing	O
Carriers	O
may	O
partner	O
with	O
vendors	O
to	O
allow	O
users	O
to	O
pay	O
for	O
services	O
by	O
SMS	O
.	O

The	O
user	O
simply	O
needs	O
to	O
text	O
a	O
prescribed	O
keyword	O
to	O
a	O
prescribed	O
number	O
(	O
shortcode	O
)	O
.	O

A	O
charge	O
is	O
then	O
added	O
to	O
the	O
user	O
’	O
s	O
bill	O
with	O
their	O
mobile	O
service	O
provider	O
.	O

Toll	O
Billing	O
Carriers	O
may	O
also	O
provide	O
payment	O
endpoints	O
over	O
a	O
web	O
page	O
.	O

The	O
user	O
visits	O
the	O
URL	O
to	O
complete	O
the	O
payment	O
and	O
enters	O
their	O
phone	O
number	O
.	O

Verification	O
that	O
the	O
request	O
is	O
coming	O
from	O
the	O
user	O
’	O
s	O
device	O
is	O
completed	O
using	O
two	O
possible	O
methods	O
:	O
The	O
user	O
connects	O
to	O
the	O
site	O
over	O
mobile	O
data	O
,	O
not	O
WiFi	O
(	O
so	O
the	O
service	O
provider	O
directly	O
handles	O
the	O
connection	O
and	O
can	O
validate	O
the	O
phone	O
number	O
)	O
;	O
or	O
The	O
user	O
must	O
retrieve	O
a	O
code	O
sent	O
to	O
them	O
via	O
SMS	O
and	O
enter	O
it	O
into	O
the	O
web	O
page	O
(	O
thereby	O
proving	O
access	O
to	O
the	O
provided	O
phone	O
number	O
)	O
.	O

Fraud	O
Both	O
of	O
the	O
billing	O
methods	O
detailed	O
above	O
provide	O
device	O
verification	O
,	O
but	O
not	O
user	O
verification	O
.	O

The	O
carrier	O
can	O
determine	O
that	O
the	O
request	O
originates	O
from	O
the	O
user	O
’	O
s	O
device	O
,	O
but	O
does	O
not	O
require	O
any	O
interaction	O
from	O
the	O
user	O
that	O
can	O
not	O
be	O
automated	O
.	O

Malware	O
authors	O
use	O
injected	O
clicks	O
,	O
custom	O
HTML	O
parsers	O
and	O
SMS	O
receivers	O
to	O
automate	O
the	O
billing	O
process	O
without	O
requiring	O
any	O
interaction	O
from	O
the	O
user	O
.	O

STRING	O
&	O
DATA	O
OBFUSCATION	O
Bread	O
apps	O
have	O
used	O
many	O
innovative	O
and	O
classic	O
techniques	O
to	O
hide	O
strings	O
from	O
analysis	O
engines	O
.	O

Here	O
are	O
some	O
highlights	O
.	O

Standard	O
Encryption	O
Frequently	O
,	O
Bread	O
apps	O
take	O
advantage	O
of	O
standard	O
crypto	O
libraries	O
in	O
`	O
java.util.crypto	B-Indicator
`	O
.	O

We	O
have	O
discovered	O
apps	O
using	O
AES	O
,	O
Blowfish	O
,	O
and	O
DES	O
as	O
well	O
as	O
combinations	O
of	O
these	O
to	O
encrypt	O
their	O
strings	O
.	O

Custom	O
Encryption	O
Other	O
variants	O
have	O
used	O
custom-implemented	O
encryption	O
algorithms	O
.	O

Some	O
common	O
techniques	O
include	O
:	O
basic	O
XOR	O
encryption	O
,	O
nested	O
XOR	O
and	O
custom	O
key-derivation	O
methods	O
.	O

Some	O
variants	O
have	O
gone	O
so	O
far	O
as	O
to	O
use	O
a	O
different	O
key	O
for	O
the	O
strings	O
of	O
each	O
class	O
.	O

Split	O
Strings	O
Encrypted	O
strings	O
can	O
be	O
a	O
signal	O
that	O
the	O
code	O
is	O
trying	O
to	O
hide	O
something	O
.	O

Bread	B-Malware
has	O
used	O
a	O
few	O
tricks	O
to	O
keep	O
strings	O
in	O
plaintext	O
while	O
preventing	O
basic	O
string	O
matching	O
.	O

Going	O
one	O
step	O
further	O
,	O
these	O
substrings	O
are	O
sometimes	O
scattered	O
throughout	O
the	O
code	O
,	O
retrieved	O
from	O
static	O
variables	O
and	O
method	O
calls	O
.	O

Various	O
versions	O
may	O
also	O
change	O
the	O
index	O
of	O
the	O
split	O
(	O
e.g	O
.	O

“	O
.clic	O
”	O
and	O
“	O
k	O
(	O
)	O
;	O
”	O
)	O
.	O

Delimiters	O
Another	O
technique	O
to	O
obfuscate	O
unencrypted	O
strings	O
uses	O
repeated	O
delimiters	O
.	O

A	O
short	O
,	O
constant	O
string	O
of	O
characters	O
is	O
inserted	O
at	O
strategic	O
points	O
to	O
break	O
up	O
keywords	O
:	O
At	O
runtime	O
,	O
the	O
delimiter	O
is	O
removed	O
before	O
using	O
the	O
string	O
:	O
API	O
OBFUSCATION	O
SMS	O
and	O
toll	O
fraud	O
generally	O
requires	O
a	O
few	O
basic	O
behaviors	O
(	O
for	O
example	O
,	O
disabling	O
WiFi	O
or	O
accessing	O
SMS	O
)	O
,	O
which	O
are	O
accessible	O
by	O
a	O
handful	O
of	O
APIs	O
.	O

Given	O
that	O
there	O
are	O
a	O
limited	O
number	O
of	O
behaviors	O
required	O
to	O
identify	O
billing	O
fraud	O
,	O
Bread	B-Malware
apps	O
have	O
had	O
to	O
try	O
a	O
wide	O
variety	O
of	O
techniques	O
to	O
mask	O
usage	O
of	O
these	O
APIs	O
.	O

Reflection	O
Most	O
methods	O
for	O
hiding	O
API	O
usage	O
tend	O
to	O
use	O
Java	O
reflection	O
in	O
some	O
way	O
.	O

In	O
some	O
samples	O
,	O
Bread	B-Malware
has	O
simply	O
directly	O
called	O
the	O
Reflect	O
API	O
on	O
strings	O
decrypted	O
at	O
runtime	O
.	O

JNI	O
Bread	B-Malware
has	O
also	O
tested	O
our	O
ability	O
to	O
analyze	O
native	O
code	O
.	O

In	O
one	O
sample	O
,	O
no	O
SMS-related	O
code	O
appears	O
in	O
the	O
DEX	O
file	O
,	O
but	O
there	O
is	O
a	O
native	O
method	O
registered	O
.	O

Two	O
strings	O
are	O
passed	O
into	O
the	O
call	O
,	O
the	O
shortcode	O
and	O
keyword	O
used	O
for	O
SMS	O
billing	O
(	O
getter	O
methods	O
renamed	O
here	O
for	O
clarity	O
)	O
.	O

In	O
the	O
native	O
library	O
,	O
it	O
stores	O
the	O
strings	O
to	O
access	O
the	O
SMS	O
API	O
.	O

The	O
nativesend	O
method	O
uses	O
the	O
Java	O
Native	O
Interface	O
(	O
JNI	O
)	O
to	O
fetch	O
and	O
call	O
the	O
Android	B-System
SMS	O
API	O
.	O

The	O
following	O
is	O
a	O
screenshot	O
from	O
IDA	O
with	O
comments	O
showing	O
the	O
strings	O
and	O
JNI	O
functions	O
.	O

WebView	O
JavaScript	O
Interface	O
Continuing	O
on	O
the	O
theme	O
of	O
cross-language	O
bridges	O
,	O
Bread	B-Malware
has	O
also	O
tried	O
out	O
some	O
obfuscation	O
methods	O
utilizing	O
JavaScript	O
in	O
WebViews	O
.	O

The	O
following	O
method	O
is	O
declared	O
in	O
the	O
DEX	O
.	O

Without	O
context	O
,	O
this	O
method	O
does	O
not	O
reveal	O
much	O
about	O
its	O
intended	O
behavior	O
,	O
and	O
there	O
are	O
no	O
calls	O
made	O
to	O
it	O
anywhere	O
in	O
the	O
DEX	O
.	O

However	O
,	O
the	O
app	O
does	O
create	O
a	O
WebView	O
and	O
registers	O
a	O
JavaScript	O
interface	O
to	O
this	O
class	O
.	O

This	O
gives	O
JavaScript	O
run	O
in	O
the	O
WebView	O
access	O
to	O
this	O
method	O
.	O

The	O
app	O
loads	O
a	O
URL	O
pointing	O
to	O
a	O
Bread-controlled	O
server	O
.	O

The	O
response	O
contains	O
some	O
basic	O
HTML	O
and	O
JavaScript	O
.	O

In	O
green	O
,	O
we	O
can	O
see	O
the	O
references	O
to	O
the	O
SMS	O
API	O
.	O

In	O
red	O
,	O
we	O
see	O
those	O
values	O
being	O
passed	O
into	O
the	O
suspicious	O
Java	O
method	O
through	O
the	O
registered	O
interface	O
.	O

Now	O
,	O
using	O
these	O
strings	O
method1	O
can	O
use	O
reflection	O
to	O
call	O
sendTextMessage	O
and	O
process	O
the	O
payment	O
.	O

PACKING	O
In	O
addition	O
to	O
implementing	O
custom	O
obfuscation	O
techniques	O
,	O
apps	O
have	O
used	O
several	O
commercially	O
available	O
packers	O
including	O
:	O
Qihoo360	B-System
,	O
AliProtect	B-System
and	O
SecShell	B-System
.	O

More	O
recently	O
,	O
we	O
have	O
seen	O
Bread-related	B-Malware
apps	O
trying	O
to	O
hide	O
malicious	O
code	O
in	O
a	O
native	O
library	O
shipped	O
with	O
the	O
APK	O
.	O

Earlier	O
this	O
year	O
,	O
we	O
discovered	O
apps	O
hiding	O
a	O
JAR	O
in	O
the	O
data	O
section	O
of	O
an	O
ELF	O
file	O
which	O
it	O
then	O
dynamically	O
loads	O
using	O
DexClassLoader	O
.	O

The	O
figure	O
below	O
shows	O
a	O
fragment	O
of	O
encrypted	O
JAR	O
stored	O
in	O
.rodata	O
section	O
of	O
a	O
shared	O
object	O
shipped	O
with	O
the	O
APK	O
as	O
well	O
as	O
the	O
XOR	O
key	O
used	O
for	O
decryption	O
.	O

After	O
we	O
blocked	O
those	O
samples	O
,	O
they	O
moved	O
a	O
significant	O
portion	O
of	O
malicious	O
functionality	O
into	O
the	O
native	O
library	O
,	O
which	O
resulted	O
in	O
a	O
rather	O
peculiar	O
back	O
and	O
forth	O
between	O
Dalvik	O
and	O
native	O
code	O
:	O
COMMAND	O
&	O
CONTROL	O
Dynamic	O
Shortcodes	O
&	O
Content	O
Early	O
versions	O
of	O
Bread	O
utilized	O
a	O
basic	O
command	O
and	O
control	O
infrastructure	O
to	O
dynamically	O
deliver	O
content	O
and	O
retrieve	O
billing	O
details	O
.	O

In	O
the	O
example	O
server	O
response	O
below	O
,	O
the	O
green	O
fields	O
show	O
text	O
to	O
be	O
shown	O
to	O
the	O
user	O
.	O

The	O
red	O
fields	O
are	O
used	O
as	O
the	O
shortcode	O
and	O
keyword	O
for	O
SMS	O
billing	O
.	O

State	O
Machines	O
Since	O
various	O
carriers	O
implement	O
the	O
billing	O
process	O
differently	O
,	O
Bread	B-Malware
has	O
developed	O
several	O
variants	O
containing	O
generalized	O
state	O
machines	O
implementing	O
all	O
possible	O
steps	O
.	O

At	O
runtime	O
,	O
the	O
apps	O
can	O
check	O
which	O
carrier	O
the	O
device	O
is	O
connected	O
to	O
and	O
fetch	O
a	O
configuration	O
object	O
from	O
the	O
command	O
and	O
control	O
server	O
.	O

The	O
configuration	O
contains	O
a	O
list	O
of	O
steps	O
to	O
execute	O
with	O
URLs	O
and	O
JavaScript	O
.	O

The	O
steps	O
implemented	O
include	O
:	O
Load	O
a	O
URL	O
in	O
a	O
WebView	O
Run	O
JavaScript	O
in	O
WebView	O
Toggle	O
WiFi	O
state	O
Toggle	O
mobile	O
data	O
state	O
Read/modify	O
SMS	O
inbox	O
Solve	O
captchas	O
Captchas	O
One	O
of	O
the	O
more	O
interesting	O
states	O
implements	O
the	O
ability	O
to	O
solve	O
basic	O
captchas	O
(	O
obscured	O
letters	O
and	O
numbers	O
)	O
.	O

First	O
,	O
the	O
app	O
creates	O
a	O
JavaScript	O
function	O
to	O
call	O
a	O
Java	O
method	O
,	O
getImageBase64	O
,	O
exposed	O
to	O
WebView	O
using	O
addJavascriptInterface	O
.	O

The	O
value	O
used	O
to	O
replace	O
GET_IMG_OBJECT	O
comes	O
from	O
the	O
JSON	O
configuration	O
.	O

The	O
app	O
then	O
uses	O
JavaScript	O
injection	O
to	O
create	O
a	O
new	O
script	O
in	O
the	O
carrier	O
’	O
s	O
web	O
page	O
to	O
run	O
the	O
new	O
function	O
.	O

The	O
base64-encoded	O
image	O
is	O
then	O
uploaded	O
to	O
an	O
image	O
recognition	O
service	O
.	O

If	O
the	O
text	O
is	O
retrieved	O
successfully	O
,	O
the	O
app	O
uses	O
JavaScript	O
injection	O
again	O
to	O
submit	O
the	O
HTML	O
form	O
with	O
the	O
captcha	O
answer	O
.	O

CLOAKING	O
Client-side	O
Carrier	O
Checks	O
In	O
our	O
basic	O
command	O
&	O
control	O
example	O
above	O
,	O
we	O
didn	O
’	O
t	O
address	O
the	O
(	O
incorrectly	O
labeled	O
)	O
“	O
imei	O
”	O
field	O
.	O

This	O
contains	O
the	O
Mobile	O
Country	O
Code	O
(	O
MCC	O
)	O
and	O
Mobile	O
Network	O
Code	O
(	O
MNC	O
)	O
values	O
that	O
the	O
billing	O
process	O
will	O
work	O
for	O
.	O

In	O
this	O
example	O
,	O
the	O
server	O
response	O
contains	O
several	O
values	O
for	O
Thai	O
carriers	O
.	O

The	O
app	O
checks	O
if	O
the	O
device	O
’	O
s	O
network	O
matches	O
one	O
of	O
those	O
provided	O
by	O
the	O
server	O
.	O

If	O
it	O
does	O
,	O
it	O
will	O
commence	O
with	O
the	O
billing	O
process	O
.	O

If	O
the	O
value	O
does	O
not	O
match	O
,	O
the	O
app	O
skips	O
the	O
“	O
disclosure	O
”	O
page	O
and	O
billing	O
process	O
and	O
brings	O
the	O
user	O
straight	O
to	O
the	O
app	O
content	O
.	O

In	O
some	O
versions	O
,	O
the	O
server	O
would	O
only	O
return	O
valid	O
responses	O
several	O
days	O
after	O
the	O
apps	O
were	O
submitted	O
.	O

Server-side	O
Carrier	O
Checks	O
In	O
the	O
JavaScript	O
bridge	O
API	O
obfuscation	O
example	O
covered	O
above	O
,	O
the	O
server	O
supplied	O
the	O
app	O
with	O
the	O
necessary	O
strings	O
to	O
complete	O
the	O
billing	O
process	O
.	O

However	O
,	O
analysts	O
may	O
not	O
always	O
see	O
the	O
indicators	O
of	O
compromise	O
in	O
the	O
server	O
’	O
s	O
response	O
.	O

In	O
this	O
example	O
,	O
the	O
requests	O
to	O
the	O
server	O
take	O
the	O
following	O
form	O
:	O
Here	O
,	O
the	O
“	O
operator	O
”	O
query	O
parameter	O
is	O
the	O
Mobile	O
Country	O
Code	O
and	O
Mobile	O
Network	O
Code	O
.	O

The	O
server	O
can	O
use	O
this	O
information	O
to	O
determine	O
if	O
the	O
user	O
’	O
s	O
carrier	O
is	O
one	O
of	O
Bread	B-Malware
’	O
s	O
targets	O
.	O

If	O
not	O
,	O
the	O
response	O
is	O
scrubbed	O
of	O
the	O
strings	O
used	O
to	O
complete	O
the	O
billing	O
fraud	O
.	O

MISLEADING	O
USERS	O
Bread	B-Malware
apps	O
sometimes	O
display	O
a	O
pop-up	O
to	O
the	O
user	O
that	O
implies	O
some	O
form	O
of	O
compliance	O
or	O
disclosure	O
,	O
showing	O
terms	O
and	O
conditions	O
or	O
a	O
confirm	O
button	O
.	O

However	O
,	O
the	O
actual	O
text	O
would	O
often	O
only	O
display	O
a	O
basic	O
welcome	O
message	O
.	O

Other	O
versions	O
included	O
all	O
the	O
pieces	O
needed	O
for	O
a	O
valid	O
disclosure	O
message	O
.	O

However	O
,	O
there	O
are	O
still	O
two	O
issues	O
here	O
:	O
The	O
numbers	O
to	O
contact	O
for	O
cancelling	O
the	O
subscription	O
are	O
not	O
real	O
The	O
billing	O
process	O
commences	O
even	O
if	O
you	O
don	O
’	O
t	O
hit	O
the	O
“	O
Confirm	O
”	O
button	O
Even	O
if	O
the	O
disclosure	O
here	O
displayed	O
accurate	O
information	O
,	O
the	O
user	O
would	O
often	O
find	O
that	O
the	O
advertised	O
functionality	O
of	O
the	O
app	O
did	O
not	O
match	O
the	O
actual	O
content	O
.	O

Bread	B-Malware
apps	O
frequently	O
contain	O
no	O
functionality	O
beyond	O
the	O
billing	O
process	O
or	O
simply	O
clone	O
content	O
from	O
other	O
popular	O
apps	O
.	O

VERSIONING	O
Bread	B-Malware
has	O
also	O
leveraged	O
an	O
abuse	O
tactic	O
unique	O
to	O
app	O
stores	O
:	O
versioning	O
.	O

Some	O
apps	O
have	O
started	O
with	O
clean	O
versions	O
,	O
in	O
an	O
attempt	O
to	O
grow	O
user	O
bases	O
and	O
build	O
the	O
developer	O
accounts	O
’	O
reputations	O
.	O

Only	O
later	O
is	O
the	O
malicious	O
code	O
introduced	O
,	O
through	O
an	O
update	O
.	O

Interestingly	O
,	O
early	O
“	O
clean	O
”	O
versions	O
contain	O
varying	O
levels	O
of	O
signals	O
that	O
the	O
updates	O
will	O
include	O
malicious	O
code	O
later	O
.	O

Some	O
are	O
first	O
uploaded	O
with	O
all	O
the	O
necessary	O
code	O
except	O
the	O
one	O
line	O
that	O
actually	O
initializes	O
the	O
billing	O
process	O
.	O

Others	O
may	O
have	O
the	O
necessary	O
permissions	O
,	O
but	O
are	O
missing	O
the	O
classes	O
containing	O
the	O
fraud	O
code	O
.	O

And	O
others	O
have	O
all	O
malicious	O
content	O
removed	O
,	O
except	O
for	O
log	O
comments	O
referencing	O
the	O
payment	O
process	O
.	O

All	O
of	O
these	O
methods	O
attempt	O
to	O
space	O
out	O
the	O
introduction	O
of	O
possible	O
signals	O
in	O
various	O
stages	O
,	O
testing	O
for	O
gaps	O
in	O
the	O
publication	O
process	O
.	O

However	O
,	O
GPP	O
does	O
not	O
treat	O
new	O
apps	O
and	O
updates	O
any	O
differently	O
from	O
an	O
analysis	O
perspective	O
.	O

FAKE	O
REVIEWS	O
When	O
early	O
versions	O
of	O
apps	O
are	O
first	O
published	O
,	O
many	O
five	O
star	O
reviews	O
appear	O
with	O
comments	O
like	O
:	O
“	O
So	O
..	O
good	O
..	O
”	O
“	O
very	O
beautiful	O
”	O
Later	O
,	O
1	O
star	O
reviews	O
from	O
real	O
users	O
start	O
appearing	O
with	O
comments	O
like	O
:	O
“	O
Deception	O
”	O
“	O
The	O
app	O
is	O
not	O
honest	O
…	O
”	O
SUMMARY	O
Sheer	O
volume	O
appears	O
to	O
be	O
the	O
preferred	O
approach	O
for	O
Bread	B-Malware
developers	O
.	O

At	O
different	O
times	O
,	O
we	O
have	O
seen	O
three	O
or	O
more	O
active	O
variants	O
using	O
different	O
approaches	O
or	O
targeting	O
different	O
carriers	O
.	O

Within	O
each	O
variant	O
,	O
the	O
malicious	O
code	O
present	O
in	O
each	O
sample	O
may	O
look	O
nearly	O
identical	O
with	O
only	O
one	O
evasion	O
technique	O
changed	O
.	O

Sample	O
1	O
may	O
use	O
AES-encrypted	B-Organization
strings	O
with	O
reflection	O
,	O
while	O
Sample	O
2	O
(	O
submitted	O
on	O
the	O
same	O
day	O
)	O
will	O
use	O
the	O
same	O
code	O
but	O
with	O
plaintext	O
strings	O
.	O

At	O
peak	O
times	O
of	O
activity	O
,	O
we	O
have	O
seen	O
up	O
to	O
23	O
different	O
apps	O
from	O
this	O
family	O
submitted	O
to	O
Play	B-System
in	O
one	O
day	O
.	O

At	O
other	O
times	O
,	O
Bread	B-Malware
appears	O
to	O
abandon	O
hope	O
of	O
making	O
a	O
variant	O
successful	O
and	O
we	O
see	O
a	O
gap	O
of	O
a	O
week	O
or	O
longer	O
before	O
the	O
next	O
variant	O
.	O

This	O
family	O
showcases	O
the	O
amount	O
of	O
resources	O
that	O
malware	O
authors	O
now	O
have	O
to	O
expend	O
.	O

Google	B-System
Play	I-System
Protect	I-System
is	O
constantly	O
updating	O
detection	O
engines	O
and	O
warning	O
users	O
of	O
malicious	O
apps	O
installed	O
on	O
their	O
device	O
.	O

SELECTED	O
SAMPLES	O
Package	O
Name	O
SHA-256	O
Digest	O
com.rabbit.artcamera	B-Indicator
18c277c7953983f45f2fe6ab4c7d872b2794c256604e43500045cb2b2084103f	B-Indicator
org.horoscope.astrology.predict	B-Indicator
6f1a1dbeb5b28c80ddc51b77a83c7a27b045309c4f1bff48aaff7d79dfd4eb26	B-Indicator
com.theforest.rotatemarswallpaper	B-Indicator
4e78a26832a0d471922eb61231bc498463337fed8874db5f70b17dd06dcb9f09	B-Indicator

com.jspany.temp	B-Indicator
0ce78efa764ce1e7fb92c4de351ec1113f3e2ca4b2932feef46d7d62d6ae87f5	B-Indicator
com.hua.ru.quan	B-Indicator
780936deb27be5dceea20a5489014236796a74cc967a12e36cb56d9b8df9bc86	B-Indicator
com.rongnea.udonood	B-Indicator
8b2271938c524dd1064e74717b82e48b778e49e26b5ac2dae8856555b5489131	B-Indicator

com.mbv.a.wp	B-Indicator
01611e16f573da2c9dbc7acdd445d84bae71fecf2927753e341d8a5652b89a68	B-Indicator
com.pho.nec.sg	B-Indicator
b4822eeb71c83e4aab5ddfecfb58459e5c5e10d382a2364da1c42621f58e119b	B-Indicator
Exobot	B-Malware
(	O
Marcher	B-Malware
)	O
-	O
Android	B-System
banking	O
Trojan	O
on	O
the	O
rise	O
February	O
2017	O
Introduction	O
The	O
past	O
months	O
many	O
different	O
banking	O
Trojans	O
for	O

the	O
Android	B-System
platform	O
have	O
received	O
media	O
attention	O
.	O

One	O
of	O
these	O
,	O
called	O
Marcher	B-Malware
(	O
aka	O
Exobot	B-Malware
)	O
,	O
seems	O
to	O
be	O
especially	O
active	O
with	O
different	O
samples	O
appearing	O
on	O
a	O
daily	O
basis	O
.	O

This	O
malware	O
variant	O
also	O
appears	O
to	O
be	O
technically	O
superior	O
to	O
many	O
other	O
banking	O
Trojans	O
being	O
able	O
to	O
use	O
its	O
overlay	O
attack	O
even	O
on	O
Android	B-System
6	I-System
,	O
which	O
has	O
technical	O
improvements	O
compared	O
to	O
the	O
previous	O
Android	B-System
versions	O
to	O
prevent	O
such	O
attacks	O
.	O

The	O
main	O
infection	O
vector	O
is	O
a	O
phishing	O
attack	O
using	O
SMS/MMS	O
.	O

The	O
social	O
engineering	O
message	O
includes	O
a	O
link	O
that	O
leads	O
to	O
a	O
fake	O
version	O
of	O
a	O
popular	O
app	O
,	O
using	O
names	O
like	O
Runtastic	B-System
,	O
WhatsApp	B-System
or	O
Netflix	B-System
.	O

On	O
installation	O
,	O
the	O
app	O
requests	O
the	O
user	O
to	O
provide	O
SMS	O
storage	O
access	O
and	O
high	O
Android	B-System
privileges	O
such	O
as	O
Device	O
Admin	O
.	O

Other	O
infection	O
vectors	O
include	O
pornographic	O
websites	O
serving	O
apps	O
called	O
Adobe	B-System
Flash	I-System
or	O
YouPorn	B-System
.	O

The	O
Marcher	B-Malware
banking	O
malware	O
uses	O
two	O
main	O
attack	O
vectors	O
.	O

The	O
first	O
attack	O
vector	O
is	O
to	O
compromise	O
the	O
out	O
of	O
band	O
authentication	O
for	O
online	O
banks	O
that	O
rely	O
on	O
SMS	O
using	O
SMS	O
forwarding	O
.	O

The	O
second	O
attack	O
vector	O
,	O
the	O
overlay	O
attack	O
,	O
shows	O
a	O
customized	O
phishing	O
window	O
whenever	O
a	O
targeted	O
application	O
is	O
started	O
on	O
the	O
device	O
.	O

The	O
overlay	O
window	O
is	O
often	O
indistinguishable	O
from	O
the	O
expected	O
screen	O
(	O
such	O
as	O
a	O
login	O
screen	O
for	O
a	O
banking	O
app	O
)	O
and	O
is	O
used	O
to	O
steal	O
the	O
victim	O
’	O
s	O
banking	O
credentials	O
.	O

The	O
target	O
list	O
and	O
bank	O
specific	O
fake	O
login	O
pages	O
can	O
be	O
dynamically	O
updated	O
via	O
their	O
C2	O
panel	O
(	O
dashboard	O
back-end	O
)	O
which	O
significantly	O
increases	O
the	O
adaptability	O
and	O
scalability	O
of	O
this	O
attack	O
.	O

In	O
addition	O
,	O
this	O
type	O
of	O
Android	B-System
banking	O
malware	O
does	O
not	O
require	O
the	O
device	O
to	O
be	O
rooted	O
or	O
the	O
app	O
to	O
have	O
any	O
specific	O
Android	B-System
permission	O
(	O
besides	O
android.permission.INTERNET	B-Indicator
to	O
retrieve	O
the	O
overlay	O
contents	O
and	O
send	O
its	O
captured	O
data	O
)	O
.	O

The	O
many	O
changes	O
we	O
see	O
in	O
the	O
way	O
the	O
attacks	O
are	O
performed	O
show	O
that	O
attackers	O
are	O
heavily	O
experimenting	O
to	O
find	O
the	O
best	O
way	O
of	O
infecting	O
a	O
mobile	O
device	O
and	O
abusing	O
existing	O
functionality	O
to	O
perform	O
successful	O
phishing	O
attacks	O
.	O

The	O
next	O
stage	O
in	O
device	O
infection	O
could	O
be	O
the	O
use	O
of	O
exploit	O
kits	O
and	O
malvertising	O
,	O
which	O
would	O
be	O
quite	O
effective	O
due	O
the	O
many	O
Android	B-Vulnerability
vulnerabilities	I-Vulnerability
and	O
consumers	O
with	O
unpatched	B-Vulnerability
devices	I-Vulnerability
.	O

In	O
addition	O
future	O
Trojans	O
could	O
leverage	O
root	O
exploits	O
to	O
make	O
them	O
almost	O
impossible	O
to	O
remove	O
and	O
give	O
malicious	O
actors	O
the	O
ability	O
to	O
hook	O
generic	O
low	O
level	O
API	O
’	O
s	O
that	O
are	O
used	O
by	O
all	O
(	O
banking	O
)	O
applications	O
,	O
just	O
like	O
the	O
attack	O
vector	O
as	O
has	O
been	O
used	O
on	O
the	O
desktop	O
platform	O
for	O
years	O
.	O

Technical	O
Analysis	O
Permissions	O
Marcher	B-Malware
’	O
s	O
APK	O
size	O
is	O
fairly	O
small	O
(	O
only	O
683KB	O
for	O
sample	O
eb8f02fc30ec49e4af1560e54b53d1a7	B-Indicator
)	O
,	O
much	O
smaller	O
than	O
most	O
legitimate	O
apps	O
and	O
other	O
popular	O
mobile	O
malware	O
samples	O
.	O

This	O
sample	O
only	O
includes	O
Dalvik	O
bytecode	O
and	O
resources	O
without	O
any	O
native	O
libraries	O
.	O

The	O
package	O
name	O
(	O
vyn.hhsdzgvoexobmkygffzwuewrbikzud	B-Indicator
)	O
and	O
its	O
many	O
activities	O
and	O
services	O
have	O
randomized	O
names	O
,	O
probably	O
to	O
make	O
it	O
a	O
bit	O
more	O
difficult	O
to	O
detect	O
the	O
package	O
using	O
blacklisting	O
.	O

The	O
set	O
of	O
permissions	O
required	O
by	O
Marcher	B-Malware
according	O
to	O
the	O
manifest	O
is	O
as	O
follows	O
:	O
∗	O
android.permission.CHANGE_NETWORK_STATE	B-Indicator
(	O
change	O
network	O
connectivity	O
state	O
)	O
∗	O
android.permission.SEND_SMS	B-Indicator
(	O
send	O
SMS	O
messages	O
)	O
∗	O
android.permission.USES_POLICY_FORCE_LOCK	B-Indicator
(	O
lock	O
the	O
device	O
)	O
∗	O
android.permission.RECEIVE_BOOT_COMPLETED	B-Indicator
(	O
start	O
malware	O
when	O
device	O
boots	O
)	O
∗	O
android.permission.INTERNET	B-Indicator
(	O
communicate	O
with	O
the	O
internet	O
)	O
∗	O
android.permission.VIBRATE	B-Indicator

(	O
control	O
the	O
vibrator	O
)	O
∗	O
android.permission.ACCESS_WIFI_STATE	B-Indicator
(	O
view	O
information	O
about	O
the	O
status	O
of	O
Wi-Fi	O
)	O
∗	O
android.permission.WRITE_SMS	B-Indicator
(	O
edit/delete	O
SMS	O
)	O
∗	O
android.permission.ACCESS_NETWORK_STATE	B-Indicator
(	O
view	O
the	O
status	O
of	O
all	O
networks	O
)	O
∗	O
android.permission.WAKE_LOCK	B-Indicator
(	O
prevent	O
the	O
phone	O
from	O
going	O
to	O
sleep	O
)	O
∗	O
android.permission.GET_TASKS	B-Indicator
(	O
retrieve	O
running	O
applications	O
)	O
∗	O
android.permission.CALL_PHONE	B-Indicator
(	O
call	O
phone	O
numbers	O
)	O

∗	O
android.permission.WRITE_SETTINGS	B-Indicator
(	O
read/write	O
global	O
system	O
settings	O
)	O
∗	O
android.permission.RECEIVE_SMS	B-Indicator
(	O
intercept	O
SMS	O
messages	O
)	O
∗	O
android.permission.READ_PHONE_STATE	B-Indicator
(	O
read	O
phone	O
details	O
of	O
the	O
device	O
such	O
as	O
phone	O
number	O
and	O
serial	O
number	O
)	O
∗	O
android.permission.CHANGE_WIFI_STATE	B-Indicator
(	O
connect	O
to	O
and	O
disconnect	O
from	O
Wi-Fi	O
networks	O
and	O
make	O
changes	O
to	O
configured	O
networks	O
)	O
∗	O
android.permission.READ_CONTACTS	B-Indicator
(	O
read	O
all	O
contact	O
data	O
)	O
*	O
android.permission.READ_SMS	B-Indicator

(	O
read	O
SMS	O
messages	O
)	O
Obviously	O
a	O
fairly	O
significant	O
list	O
of	O
permissions	O
of	O
which	O
many	O
are	O
suspicious	O
,	O
especially	O
when	O
combined	O
.	O

Runtastic	B-System
sample	O
permission	O
prompt	O
Runtastic	B-System
sample	O
permission	O
prompt	O
Checking	O
foreground	O
app	O
Marcher	B-Malware
is	O
one	O
of	O
the	O
few	O
Android	O
banking	O
Trojans	O
to	O
use	O
the	O
AndroidProcesses	O
library	O
,	O
which	O
enables	O
the	O
application	O
to	O
obtain	O
the	O
name	O
of	O
the	O
Android	B-System
package	O
that	O
is	O
currently	O
running	O
in	O
the	O
foreground	O
.	O

This	O
library	O
is	O
used	O
because	O
it	O
uses	O
the	O
only	O
(	O
publicly	O
known	O
)	O
way	O
to	O
retrieve	O
this	O
information	O
on	O
Android	B-System
6	I-System
(	O
using	O
the	O
process	O
OOM	O
score	O
read	O
from	O
the	O
/proc	O
directory	O
)	O
.	O

When	O
the	O
current	O
app	O
on	O
the	O
foreground	O
matches	O
with	O
an	O
app	O
targeted	O
by	O
the	O
malware	O
,	O
the	O
Trojan	O
will	O
show	O
the	O
corresponding	O
phishing	O
overlay	O
,	O
making	O
the	O
user	O
think	O
it	O
is	O
the	O
app	O
that	O
was	O
just	O
started	O
.	O

Dynamic	O
overlays	O
When	O
victims	O
open	O
up	O
a	O
targeted	O
app	O
,	O
Marcher	B-Malware
smoothly	O
displays	O
an	O
overlay	O
,	O
a	O
customized	O
WebView	O
,	O
looks	O
in	O
its	O
application	O
preferences	O
(	O
main_prefs.xml	O
)	O
and	O
decides	O
which	O
specified	O
URL	O
is	O
needed	O
for	O
the	O
targeted	O
app	O
.	O

The	O
complete	O
list	O
of	O
apps	O
can	O
be	O
seen	O
below	O
.	O

The	O
phishing	O
pages	O
shown	O
in	O
the	O
overlay	O
use	O
Ajax	O
calls	O
to	O
communicate	O
with	O
a	O
PHP	O
back-end	O
which	O
stores	O
all	O
user	O
input	O
.	O

The	O
C2	O
backend	O
url	O
looks	O
like	O
this	O
:	O
https	B-Indicator
:	I-Indicator
//evilhost/c2folder/njs2/	I-Indicator
?	I-Indicator

fields	I-Indicator
[	I-Indicator
]	I-Indicator
.	O

There	O
is	O
no	O
way	O
to	O
access	O
the	O
original	O
app	O
again	O
even	O
if	O
victims	O
terminate	O
the	O
overlay	O
process	O
and	O
reopen	O
app	O
,	O
until	O
credit	O
card	O
(	O
name	O
,	O
number	O
,	O
expiry	O
date	O
,	O
security	O
code	O
)	O
and/or	O
bank	O
information	O
(	O
PIN	O
,	O
VBV	O
passcode	O
,	O
date	O
of	O
birth	O
,	O
etc	O
.	O

)	O
are	O
filled	O
in	O
and	O
verified	O
.	O

The	O
information	O
is	O
then	O
stored	O
in	O
local	O
app	O
database	O
as	O
well	O
as	O
sent	O
to	O
the	O
backend	O
.	O

Agent	B-Malware
Smith	I-Malware
:	O
A	O
New	O
Species	O
of	O
Mobile	O
Malware	O
July	O
10	O
,	O
2019	O
Check	B-Organization
Point	I-Organization
Researchers	O
recently	O
discovered	O
a	O
new	O
variant	O
of	O
mobile	O
malware	O
that	O
quietly	O
infected	O
around	O
25	O
million	O
devices	O
,	O
while	O
the	O
user	O
remains	O
completely	O
unaware	O
.	O

Disguised	O
as	O
Google	B-Organization
related	O
app	O
,	O
the	O
core	O
part	O
of	O
malware	O
exploits	O
various	O
known	O
Android	B-Vulnerability
vulnerabilities	I-Vulnerability
and	O
automatically	O
replaces	O
installed	O
apps	O
on	O
the	O
device	O
with	O
malicious	O
versions	O
without	O
the	O
user	O
’	O
s	O
interaction	O
.	O

This	O
unique	O
on-device	O
,	O
just-in-time	O
(	O
JIT	O
)	O
approach	O
inspired	O
researchers	O
to	O
dub	O
this	O
malware	O
as	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
.	O

“	O
Agent	B-Malware
Smith	I-Malware
”	O
currently	O
uses	O
its	O
broad	O
access	O
to	O
the	O
device	O
’	O
s	O
resources	O
to	O
show	O
fraudulent	O
ads	O
for	O
financial	O
gain	O
.	O

This	O
activity	O
resembles	O
previous	O
campaigns	O
such	O
as	O
Gooligan	B-Malware
,	O
HummingBad	B-Malware
and	O
CopyCat	B-Malware
.	O

The	O
primary	O
targets	O
,	O
so	O
far	O
,	O
are	O
based	O
in	O
India	O
though	O
other	O
Asian	O
countries	O
such	O
as	O
Pakistan	O
and	O
Bangladesh	O
are	O
also	O
affected	O
.	O

In	O
a	O
much-improved	O
Android	B-System
security	O
environment	O
,	O
the	O
actors	O
behind	O
Agent	B-Malware
Smith	I-Malware
seem	O
to	O
have	O
moved	O
into	O
the	O
more	O
complex	O
world	O
of	O
constantly	O
searching	O
for	O
new	O
loopholes	O
,	O
such	O
as	O
Janus	B-Vulnerability
,	O
Bundle	B-Vulnerability
and	O
Man-in-the-Disk	B-Vulnerability
,	O
to	O
achieve	O
a	O
3-stage	O
infection	O
chain	O
,	O
in	O
order	O
to	O
build	O
a	O
botnet	O
of	O
controlled	O
devices	O
to	O
earn	O
profit	O
for	O
the	O
perpetrator	O
.	O

“	O
Agent	B-Malware
Smith	I-Malware
”	O
is	O
possibly	O
the	O
first	O
campaign	O
seen	O
that	O
ingrates	O
and	O
weaponized	O
all	O
these	O
loopholes	O
and	O
are	O
described	O
in	O
detail	O
below	O
.	O

In	O
this	O
case	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
is	O
being	O
used	O
to	O
for	O
financial	O
gain	O
through	O
the	O
use	O
of	O
malicious	O
advertisements	O
.	O

However	O
,	O
it	O
could	O
easily	O
be	O
used	O
for	O
far	O
more	O
intrusive	O
and	O
harmful	O
purposes	O
such	O
as	O
banking	O
credential	O
theft	O
.	O

Indeed	O
,	O
due	O
to	O
its	O
ability	O
to	O
hide	O
it	O
’	O
s	O
icon	O
from	O
the	O
launcher	O
and	O
impersonates	O
any	O
popular	O
existing	O
apps	O
on	O
a	O
device	O
,	O
there	O
are	O
endless	O
possibilities	O
for	O
this	O
sort	O
of	O
malware	O
to	O
harm	O
a	O
user	O
’	O
s	O
device	O
.	O

Check	B-Organization
Point	I-Organization
Research	O
has	O
submitted	O
data	O
to	O
Google	B-Organization
and	O
law	O
enforcement	O
units	O
to	O
facilitate	O
further	O
investigation	O
.	O

As	O
a	O
result	O
,	O
information	O
related	O
to	O
the	O
malicious	O
actor	O
is	O
tentatively	O
redacted	O
in	O
this	O
publication	O
.	O

Check	B-Organization
Point	I-Organization
has	O
worked	O
closely	O
with	O
Google	B-Organization
and	O
at	O
the	O
time	O
of	O
publishing	O
,	O
no	O
malicious	O
apps	O
remain	O
on	O
the	O
Play	B-System
Store	I-System
.	O

Encounter	O
In	O
early	O
2019	O
,	O
the	O
Check	B-Organization
Point	I-Organization
Research	O
team	O
observed	O
a	O
surge	O
of	O
Android	B-System
malware	O
attack	O
attempts	O
against	O
users	O
in	O
India	O
which	O
had	O
strong	O
characteristics	O
of	O
Janus	B-Vulnerability
vulnerability	O
abuse	O
;	O
All	O
samples	O
our	O
team	O
collected	O
during	O
preliminary	O
investigation	O
had	O
the	O
ability	O
to	O
hide	O
their	O
app	O
icons	O
and	O
claim	O
to	O
be	O
Google	B-Organization
related	O
updaters	O
or	O
vending	O
modules	O
(	O
a	O
key	O
component	O
of	O
Google	B-System
Play	I-System
framework	O
)	O
.	O

Upon	O
further	O
analysis	O
it	O
became	O
clear	O
this	O
application	O
was	O
as	O
malicious	O
as	O
they	O
come	O
and	O
initially	O
resembled	O
the	O
CopyCat	B-Malware
malware	O
,	O
discovered	O
by	O
Check	B-Organization
Point	I-Organization
Research	O
back	O
in	O
April	O
2016	O
.	O

As	O
the	O
research	O
progressed	O
,	O
it	O
started	O
to	O
reveal	O
unique	O
characteristics	O
which	O
made	O
us	O
believe	O
we	O
were	O
looking	O
at	O
an	O
all-new	O
malware	O
campaign	O
found	O
in	O
the	O
wild	O
.	O

After	O
a	O
series	O
of	O
technical	O
analysis	O
(	O
which	O
is	O
covered	O
in	O
detail	O
below	O
)	O
and	O
heuristic	O
threat	O
hunting	O
,	O
we	O
discovered	O
that	O
a	O
complete	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
infection	O
has	O
three	O
main	O
phases	O
:	O
A	O
dropper	O
app	O
lures	O
victim	O
to	O
install	O
itself	O
voluntarily	O
.	O

The	O
initial	O
dropper	O
has	O
a	O
weaponized	O
Feng	O
Shui	O
Bundle	O
as	O
encrypted	O
asset	O
files	O
.	O

Dropper	O
variants	O
are	O
usually	O
barely	O
functioning	O
photo	O
utility	O
,	O
games	O
,	O
or	O
sex	O
related	O
apps	O
.	O

The	O
dropper	O
automatically	O
decrypts	O
and	O
installs	O
its	O
core	O
malware	O
APK	O
which	O
later	O
conducts	O
malicious	O
patching	O
and	O
app	O
updates	O
.	O

The	O
core	O
malware	O
is	O
usually	O
disguised	O
as	O
Google	B-Organization
Updater	O
,	O
Google	B-Organization
Update	O
for	O
U	O
or	O
“	O
com.google.vending	B-Indicator
”	O
.	O

The	O
core	O
malware	O
’	O
s	O
icon	O
is	O
hidden	O
.	O

The	O
core	O
malware	O
extracts	O
the	O
device	O
’	O
s	O
installed	O
app	O
list	O
.	O

If	O
it	O
finds	O
apps	O
on	O
its	O
prey	O
list	O
(	O
hard-coded	O
or	O
sent	O
from	O
C	O
&	O
C	O
server	O
)	O
,	O
it	O
will	O
extract	O
the	O
base	O
APK	O
of	O
the	O
target	O
innocent	O
app	O
on	O
the	O
device	O
,	O
patch	O
the	O
APK	O
with	O
malicious	O
ads	O
modules	O
,	O
install	O
the	O
APK	O
back	O
and	O
replace	O
the	O
original	O
one	O
as	O
if	O
it	O
is	O
an	O
update	O
.	O

“	O
Agent	B-Malware
Smith	I-Malware
”	O
repacks	O
its	O
prey	O
apps	O
at	O
smali/baksmali	O
code	O
level	O
.	O

During	O
the	O
final	O
update	O
installation	O
process	O
,	O
it	O
relies	O
on	O
the	O
Janus	B-Vulnerability
vulnerability	O
to	O
bypass	O
Android	B-System
’	O
s	O
APK	O
integrity	O
checks	O
.	O

Upon	O
kill	O
chain	O
completion	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
will	O
then	O
hijack	O
compromised	O
user	O
apps	O
to	O
show	O
ads	O
.	O

In	O
certain	O
situations	O
,	O
variants	O
intercept	O
compromised	O
apps	O
’	O
original	O
legitimate	O
ads	O
display	O
events	O
and	O
report	O
back	O
to	O
the	O
intended	O
ad-exchange	O
with	O
the	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
hacker	O
’	O
s	O
ad	O
IDs	O
.	O

Our	O
intelligence	O
shows	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
droppers	O
proliferate	O
through	O
third-party	O
app	O
store	O
“	O
9Apps	B-System
”	O
,	O
a	O
UC	O
team	O
backed	O
store	O
,	O
targeted	O
mostly	O
at	O
Indian	O
(	O
Hindi	O
)	O
,	O
Arabic	O
,	O
and	O
Indonesian	O
users	O
.	O

“	O
Agent	B-Malware
Smith	I-Malware
”	O
itself	O
,	O
though	O
,	O
seems	O
to	O
target	O
mainly	O
India	O
users	O
.	O

Unlike	O
previously	O
discovered	O
non	O
Google	B-System
Play	I-System
centric	O
campaigns	O
whose	O
victims	O
almost	O
exclusively	O
come	O
from	O
less	O
developed	O
countries	O
and	O
regions	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
successfully	O
penetrated	O
into	O
noticeable	O
number	O
of	O
devices	O
in	O
developed	O
countries	O
such	O
as	O
Saudi	O
Arabia	O
,	O
UK	O
and	O
US	O
.	O

Technical	O
Analysis	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
has	O
a	O
modular	O
structure	O
and	O
consists	O
of	O
the	O
following	O
modules	O
:	O
Loader	O
Core	O
Boot	O
Patch	O
AdSDK	O
Updater	O
As	O
stated	O
above	O
,	O
the	O
first	O
step	O
of	O
this	O
infection	O
chain	O
is	O
the	O
dropper	O
.	O

The	O
dropper	O
is	O
a	O
repacked	O
legitimate	O
application	O
which	O
contains	O
an	O
additional	O
piece	O
of	O
code	O
–	O
“	O
loader	O
”	O
.	O

The	O
loader	O
has	O
a	O
very	O
simple	O
purpose	O
,	O
extract	O
and	O
run	O
the	O
“	O
core	O
”	O
module	O
of	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
.	O

The	O
“	O
core	O
”	O
module	O
communicates	O
with	O
the	O
C	O
&	O
C	O
server	O
,	O
receiving	O
the	O
predetermined	O
list	O
of	O
popular	O
apps	O
to	O
scan	O
the	O
device	O
for	O
.	O

If	O
any	O
application	O
from	O
that	O
list	O
was	O
found	O
,	O
it	O
utilizes	O
the	O
Janus	B-Vulnerability
vulnerability	O
to	O
inject	O
the	O
“	O
boot	O
”	O
module	O
into	O
the	O
repacked	O
application	O
.	O

After	O
the	O
next	O
run	O
of	O
the	O
infected	O
application	O
,	O
the	O
“	O
boot	O
”	O
module	O
will	O
run	O
the	O
“	O
patch	O
”	O
module	O
,	O
which	O
hooks	O
the	O
methods	O
from	O
known	O
ad	O
SDKs	O
to	O
its	O
own	O
implementation	O
.	O

Figure	O
1	O
:	O
‘	O
Agent	B-Malware
Smith	I-Malware
’	O
s	O
modular	O
structure	O
Technical	O
Analysis	O
–	O
Loader	O
Module	O
The	O
“	O
loader	O
”	O
module	O
,	O
as	O
stated	O
above	O
,	O
extracts	O
and	O
runs	O
the	O
“	O
core	O
”	O
module	O
.	O

While	O
the	O
“	O
core	O
”	O
module	O
resides	O
inside	O
the	O
APK	O
file	O
,	O
it	O
is	O
encrypted	O
and	O
disguised	O
as	O
a	O
JPG	O
file	O
–	O
the	O
first	O
two	O
bytes	O
are	O
actually	O
the	O
magic	O
header	O
of	O
JPG	O
files	O
,	O
while	O
the	O
rest	O
of	O
the	O
data	O
is	O
encoded	O
with	O
an	O
XOR	O
cipher	O
.	O

Figure	O
2	O
:	O
“	O
Agent	B-Malware
Smith	I-Malware
’	O
s	O
jpg	O
file	O
structure	O
After	O
the	O
extraction	O
,	O
the	O
“	O
loader	O
”	O
module	O
adds	O
the	O
code	O
to	O
the	O
application	O
while	O
using	O
the	O
legitimate	O
mechanism	O
by	O
Android	B-System
to	O
handle	O
large	O
DEX	O
files	O
.	O

Figure	O
3	O
:	O
Loading	O
core	O
malicious	O
code	O
into	O
the	O
benign	O
application	O
Once	O
the	O
“	O
core	O
”	O
module	O
is	O
extracted	O
and	O
loaded	O
,	O
the	O
“	O
loader	O
”	O
uses	O
the	O
reflection	O
technique	O
to	O
initialize	O
and	O
start	O
the	O
“	O
core	O
”	O
module	O
.	O

Figure	O
4	O
:	O
Loader	O
calls	O
initialization	O
method	O
Technical	O
Analysis	O
–	O
Core	O
Module	O
With	O
the	O
main	O
purpose	O
of	O
spreading	O
the	O
infection	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
implements	O
in	O
the	O
“	O
core	O
”	O
module	O
:	O
A	O
series	O
of	O
‘	O
Bundle	B-Vulnerability
’	O
vulnerabilities	O
,	O
which	O
is	O
used	O
to	O
install	O
applications	O
without	O
the	O
victim	O
’	O
s	O
awareness	O
.	O

The	O
Janus	B-Vulnerability
vulnerability	O
,	O
which	O
allows	O
the	O
actor	O
to	O
replace	O
any	O
application	O
with	O
an	O
infected	O
version	O
.	O

The	O
“	O
core	O
”	O
module	O
contacts	O
the	O
C	O
&	O
C	O
server	O
,	O
trying	O
to	O
get	O
a	O
fresh	O
list	O
of	O
applications	O
to	O
search	O
for	O
,	O
or	O
if	O
that	O
fails	O
,	O
use	O
a	O
default	O
app	O
list	O
:	O
whatsapp	B-System
lenovo.anyshare.gps	B-Indicator
mxtech.videoplayer.ad	B-Indicator
jio.jioplay.tv	B-Indicator
jio.media.jiobeats	B-Indicator
jiochat.jiochatapp	B-Indicator
jio.join	B-Indicator
good.gamecollection	B-Indicator
opera.mini.native	B-Indicator
startv.hotstar	B-Indicator
meitu.beautyplusme	B-Indicator
domobile.applock	B-Indicator
touchtype.swiftkey	B-Indicator
flipkart.android	B-Indicator
cn.xender	B-Indicator

eterno	O
truecaller	O
For	O
each	O
application	O
on	O
the	O
list	O
,	O
the	O
“	O
core	O
”	O
module	O
checks	O
for	O
a	O
matching	O
version	O
and	O
MD5	O
hash	O
of	O
the	O
installed	O
application	O
,	O
and	O
also	O
checks	O
for	O
the	O
application	O
running	O
in	O
the	O
user-space	O
.	O

If	O
all	O
conditions	O
are	O
met	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
tries	O
to	O
infect	O
the	O
application	O
.	O

The	O
“	O
core	O
”	O
module	O
will	O
use	O
one	O
of	O
two	O
methods	O
to	O
infect	O
the	O
application	O
–	O
Decompile	O
and	O
Binary	O
.	O

The	O
decompile	O
method	O
is	O
based	O
on	O
the	O
fact	O
that	O
Android	B-System
applications	O
are	O
Java-based	O
,	O
meaning	O
it	O
is	O
possible	O
to	O
recompile	O
it	O
.	O

Therefore	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
decompiles	O
both	O
the	O
original	O
application	O
and	O
the	O
malicious	O
payload	O
and	O
fuses	O
them	O
together	O
.	O

Figure	O
5	O
:	O
core	O
module	O
mixes	O
malicious	O
payload	O
with	O
the	O
original	O
application	O
While	O
decompiling	O
the	O
original	O
app	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
has	O
the	O
opportunity	O
to	O
modify	O
the	O
methods	O
inside	O
,	O
replace	O
some	O
of	O
the	O
methods	O
in	O
the	O
original	O
application	O
that	O
handles	O
advertisement	O
with	O
its	O
own	O
code	O
and	O
focus	O
on	O
methods	O
communicating	O
with	O
‘	O
AdMob	B-System
’	O
,	O
‘	O
Facebook	B-System
’	O
,	O
‘	O
MoPub	B-System
’	O
and	O
‘	O
Unity	B-System
Ads	I-System
’	O
.	O

Figure	O
6	O
:	O
Targeted	O
ad	O
network	O
Figure	O
7	O
:	O
Injection	O
example	O
After	O
all	O
of	O
the	O
required	O
changes	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
compiles	O
the	O
application	O
and	O
builds	O
a	O
DEX	O
file	O
containing	O
both	O
the	O
original	O
code	O
of	O
the	O
original	O
application	O
and	O
the	O
malicious	O
payload	O
.	O

In	O
some	O
cases	O
,	O
the	O
decompilation	O
process	O
will	O
fail	O
,	O
and	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
will	O
try	O
another	O
method	O
for	O
infecting	O
the	O
original	O
application	O
–	O
A	O
binary	O
patch	O
,	O
which	O
simply	O
provides	O
a	O
binary	O
file	O
of	O
the	O
“	O
boot	O
”	O
module	O
of	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
.	O

Once	O
the	O
payload	O
is	O
prepared	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
uses	O
it	O
to	O
build	O
another	O
APK	O
file	O
,	O
exploiting	O
the	O
Janus	B-Vulnerability
vulnerability	O
:	O
Figure	O
8	O
:	O
The	O
new	O
infected	O
APK	O
file	O
structure	O
Solely	O
injecting	O
the	O
code	O
of	O
the	O
loader	O
is	O
not	O
enough	O
.	O

As	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
uses	O
a	O
modular	O
approach	O
,	O
and	O
as	O
stated	O
earlier	O
,	O
the	O
original	O
loader	O
extracts	O
everything	O
from	O
the	O
assets	O
,	O
the	O
usage	O
of	O
the	O
Janus	B-Vulnerability
vulnerability	O
can	O
only	O
change	O
the	O
code	O
of	O
the	O
original	O
application	O
,	O
not	O
the	O
resources	O
.	O

This	O
means	O
that	O
the	O
only	O
thing	O
possible	O
in	O
this	O
case	O
is	O
to	O
replace	O
its	O
DEX	O
file	O
.	O

To	O
overcome	O
this	O
issue	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
found	O
another	O
solution	O
.	O

Seeing	O
as	O
the	O
system	O
loader	O
of	O
the	O
DEX	O
files	O
(	O
ART	O
)	O
fully	O
ignores	O
everything	O
that	O
goes	O
after	O
the	O
data	O
section	O
,	O
the	O
patcher	O
writes	O
all	O
of	O
its	O
resources	O
right	O
there	O
.	O

This	O
action	O
changes	O
the	O
original	O
file	O
size	O
of	O
the	O
DEX	O
file	O
,	O
which	O
makes	O
the	O
malicious	O
resources	O
a	O
part	O
of	O
the	O
DEX	O
file	O
,	O
a	O
section	O
that	O
is	O
ignored	O
by	O
the	O
signature	O
validation	O
process	O
.	O

Figure	O
9	O
:	O
Malware	O
secretly	O
adds	O
malicious	O
resources	O
to	O
the	O
DEX	O
file	O
Now	O
,	O
after	O
the	O
alteration	O
of	O
the	O
original	O
application	O
,	O
Android	B-System
’	O
s	O
package	O
manager	O
will	O
think	O
that	O
this	O
is	O
an	O
update	O
for	O
the	O
application	O
signed	O
by	O
the	O
same	O
certificate	O
,	O
but	O
in	O
reality	O
,	O
it	O
will	O
execute	O
the	O
malicious	O
DEX	O
file	O
.	O

Even	O
now	O
,	O
this	O
is	O
still	O
not	O
enough	O
.	O

“	O
Agent	B-Malware
Smith	I-Malware
”	O
needs	O
to	O
be	O
updated/installed	O
without	O
the	O
user	O
’	O
s	O
consent	O
.	O

To	O
achieve	O
this	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
utilizes	O
a	O
series	O
of	O
1-day	B-Vulnerability
vulnerabilities	I-Vulnerability
,	O
which	O
allows	O
any	O
application	O
to	O
run	O
an	O
activity	O
inside	O
a	O
system	O
application	O
,	O
even	O
if	O
this	O
activity	O
is	O
not	O
exported	O
.	O

The	O
malicious	O
application	O
sends	O
a	O
request	O
to	O
choose	O
a	O
network	O
account	O
,	O
a	O
specific	O
account	O
that	O
can	O
only	O
be	O
processed	O
by	O
authentication	O
services	O
exported	O
by	O
the	O
malicious	O
application	O
.	O

The	O
system	O
service	O
‘	O
AccountManagerService	O
’	O
looks	O
for	O
the	O
application	O
that	O
can	O
process	O
this	O
request	O
.	O

While	O
doing	O
so	O
,	O
it	O
will	O
reach	O
a	O
service	O
exported	O
by	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
,	O
and	O
sends	O
out	O
an	O
authentication	O
request	O
that	O
would	O
lead	O
to	O
a	O
call	O
to	O
the	O
‘	O
addAccount	O
’	O
method	O
.	O

Then	O
,	O
a	O
request	O
is	O
formed	O
in	O
such	O
a	O
way	O
that	O
an	O
activity	O
that	O
installs	O
the	O
application	O
is	O
called	O
,	O
bypassing	O
all	O
security	O
checks	O
.	O

Figure	O
10	O
:	O
The	O
algorithm	O
of	O
the	O
malicious	O
update	O
,	O
while	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
updates	O
application	O
If	O
all	O
that	O
has	O
failed	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
turns	O
to	O
Man-in-the-Disk	B-Vulnerability
vulnerability	O
for	O
‘	O
SHAREit	B-System
’	O
or	O
‘	O
Xender	B-System
’	O
applications	O
.	O

This	O
is	O
a	O
very	O
simple	O
process	O
,	O
which	O
is	O
replacing	O
their	O
update	O
file	O
on	O
SD	O
card	O
with	O
its	O
own	O
malicious	O
payload	O
.	O

Figure	O
11	O
:	O
‘	O
Agent	B-Malware
Smith	I-Malware
’	O
uses	O
man-in-disk	B-Vulnerability
to	O
install	O
the	O
malicious	O
update	O
Technical	O
Analysis	O
–	O
Boot	O
Module	O
The	O
“	O
boot	O
”	O
module	O
is	O
basically	O
another	O
“	O
loader	O
”	O
module	O
,	O
but	O
this	O
time	O
it	O
’	O
s	O
executed	O
in	O
the	O
infected	O
application	O
.	O

The	O
purpose	O
of	O
this	O
module	O
is	O
to	O
extract	O
and	O
execute	O
a	O
malicious	O
payload	O
–	O
the	O
“	O
patch	O
”	O
module	O
.	O

The	O
infected	O
application	O
contains	O
its	O
payload	O
inside	O
the	O
DEX	O
file	O
.	O

All	O
that	O
is	O
needed	O
is	O
to	O
get	O
the	O
original	O
size	O
of	O
the	O
DEX	O
file	O
and	O
read	O
everything	O
that	O
comes	O
after	O
this	O
offset	O
.	O

Figure	O
12	O
:	O
Boot	O
module	O
After	O
the	O
patch	O
module	O
is	O
extracted	O
,	O
the	O
“	O
boot	O
”	O
module	O
executes	O
it	O
,	O
using	O
the	O
same	O
method	O
described	O
in	O
the	O
“	O
loader	O
”	O
module	O
.	O

The	O
“	O
boot	O
”	O
module	O
has	O
placeholder	O
classes	O
for	O
the	O
entry	O
points	O
of	O
the	O
infected	O
applications	O
.	O

This	O
allows	O
the	O
“	O
boot	O
”	O
module	O
to	O
execute	O
the	O
payloads	O
when	O
the	O
infected	O
application	O
is	O
started	O
.	O

Figure	O
13	O
:	O
placeholder	O
classes	O
in	O
Boot	O
module	O
Technical	O
Analysis	O
–	O
Patch	O
Module	O
When	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
has	O
reached	O
its	O
goal	O
–	O
a	O
malicious	O
payload	O
running	O
inside	O
the	O
original	O
application	O
,	O
with	O
hooks	O
on	O
various	O
methods	O
–	O
at	O
this	O
point	O
,	O
everything	O
lies	O
with	O
maintaining	O
the	O
required	O
code	O
in	O
case	O
of	O
an	O
update	O
for	O
the	O
original	O
application	O
.	O

While	O
investing	O
a	O
lot	O
of	O
resources	O
in	O
the	O
development	O
of	O
this	O
malware	O
,	O
the	O
actor	O
behind	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
does	O
not	O
want	O
a	O
real	O
update	O
to	O
remove	O
all	O
of	O
the	O
changes	O
made	O
,	O
so	O
here	O
is	O
where	O
the	O
“	O
patch	O
”	O
module	O
comes	O
in	O
to	O
play	O
With	O
the	O
sole	O
purpose	O
of	O
disabling	O
automatic	O
updates	O
for	O
the	O
infected	O
application	O
,	O
this	O
module	O
observes	O
the	O
update	O
directory	O
for	O
the	O
original	O
application	O
and	O
removes	O
the	O
file	O
once	O
it	O
appears	O
.	O

Another	O
trick	O
in	O
“	O
Agent	B-Malware
Smith	I-Malware
’	O
s	O
arsenal	O
is	O
to	O
change	O
the	O
settings	O
of	O
the	O
update	O
timeout	O
,	O
making	O
the	O
original	O
application	O
wait	O
endlessly	O
for	O
the	O
update	O
check	O
.	O

Figure	O
14	O
:	O
disabling	O
infected	O
apps	O
auto-update	O
Figure	O
15	O
:	O
changing	O
the	O
settings	O
of	O
the	O
update	O
timeout	O
The	O
Ad	O
Displaying	O
Payload	O
Following	O
all	O
of	O
the	O
above	O
,	O
now	O
is	O
the	O
time	O
to	O
take	O
a	O
look	O
into	O
the	O
actual	O
payload	O
that	O
displays	O
ads	O
to	O
the	O
victim	O
.	O

In	O
the	O
injected	O
payload	O
,	O
the	O
module	O
implements	O
the	O
method	O
‘	O
callActivityOnCreate	O
’	O
.	O

At	O
any	O
time	O
an	O
infected	O
application	O
will	O
create	O
an	O
activity	O
,	O
this	O
method	O
will	O
be	O
called	O
,	O
and	O
call	O
‘	O
requestAd	O
’	O
from	O
“	O
Agent	B-Malware
Smith	I-Malware
’	O
s	O
code	O
.	O

“	O
Agent	B-Malware
Smith	I-Malware
”	O
will	O
replace	O
the	O
original	O
application	O
’	O
s	O
activities	O
with	O
an	O
in-house	O
SDK	O
’	O
s	O
activity	O
,	O
which	O
will	O
show	O
the	O
banner	O
received	O
from	O
the	O
server	O
.	O

In	O
the	O
case	O
of	O
the	O
infected	O
application	O
not	O
specified	O
in	O
the	O
code	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
will	O
simply	O
show	O
ads	O
on	O
the	O
activity	O
being	O
loaded	O
.	O

Figure	O
16	O
:	O
integrating	O
an	O
in-house	O
ad	O
SDK	O
Figure	O
17	O
:	O
replacing	O
original	O
app	O
activities	O
with	O
the	O
malicious	O
ad	O
SDK	O
activity	O
Figure	O
18	O
:	O
the	O
malware	O
showing	O
ads	O
on	O
any	O
activity	O
being	O
loaded	O
Connecting	O
the	O
Dots	O
As	O
our	O
malware	O
sample	O
analysis	O
took	O
the	O
team	O
closer	O
to	O
reveal	O
the	O
“	O
Agent	O
Smith	O
”	O
campaign	O
in	O
its	O
entirety	O
and	O
it	O
is	O
here	O
that	O
the	O
C	O
&	O
C	O
server	O
investigation	O
enters	O
the	O
center	O
stage	O
.	O

We	O
started	O
with	O
most	O
frequently	O
used	O
C	O
&	O
C	O
domains	O
“	O
a	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
d.com	I-Indicator
”	O
,	O
“	O
a	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
d.net	I-Indicator
”	O
,	O
and	O
“	O
a	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
d.org	I-Indicator
”	O
.	O

Among	O
multiple	O
sub-domains	O
,	O
“	O
ad.a	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
d.org	I-Indicator
”	O
and	O
“	O
gd.a	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
d.org	I-Indicator
”	O
both	O
historically	O
resolved	O
to	O
the	O
same	O
suspicious	O
IP	O
address	O
.	O

The	O
reverse	O
DNS	O
history	O
of	O
this	O
IP	O
brought	O
“	O
ads.i	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
e.com	I-Indicator
”	O
into	O
our	O
attention	O
.	O

An	O
extended	O
malware	O
hunting	O
process	O
returned	O
to	O
us	O
a	O
large	O
set	O
of	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
dropper	O
variants	O
which	O
helped	O
us	O
further	O
deduce	O
a	O
relation	O
among	O
multiple	O
C	O
&	O
C	O
server	O
infrastructures	O
.	O

In	O
a	O
different	O
period	O
of	O
the	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
,	O
droppers	O
and	O
core	O
modules	O
used	O
various	O
combinations	O
of	O
the	O
“	O
a	O
*	O
*	O
*	O
d	O
”	O
and	O
“	O
i	O
*	O
*	O
*	O
e	O
”	O
domains	O
for	O
malicious	O
operations	O
such	O
as	O
prey	O
list	O
query	O
,	O
patch	O
request	O
and	O
ads	O
request	O
.	O

With	O
a	O
bit	O
of	O
luck	O
,	O
we	O
managed	O
to	O
find	O
logs	O
in	O
which	O
the	O
evidence	O
showed	O
“	O
Agent	B-Malware
Smith	I-Malware
’	O
s	O
C	O
&	O
C	O
front	O
end	O
routinely	O
distributes	O
a	O
workload	O
between	O
“	O
w.h	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
g.com	I-Indicator
”	O
and	O
“	O
tt.a	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
d.net	I-Indicator
”	O
.	O

An	O
in-depth	O
understanding	O
of	O
the	O
“	O
Agent	B-Malware
Smith	I-Malware
’	O
s	O
campaign	O
C	O
&	O
C	O
infrastructure	O
enabled	O
us	O
to	O
reach	O
the	O
conclusion	O
that	O
the	O
owner	O
of	O
“	B-Indicator
i	I-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
e.com	I-Indicator
”	O
,	O
“	O
h	B-Indicator
*	I-Indicator
*	I-Indicator
*	I-Indicator
g.com	I-Indicator
”	O
is	O
the	O
group	O
of	O
hackers	O
behind	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
.	O

Figure	O
19	O
:	O
C	O
&	O
C	O
infrastructure	O
diagram	O
The	O
Infection	O
Landscape	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
droppers	O
show	O
a	O
very	O
greedy	O
infection	O
tactic	O
.	O

It	O
’	O
s	O
not	O
enough	O
for	O
this	O
malware	O
family	O
to	O
swap	O
just	O
one	O
innocent	O
application	O
with	O
an	O
infected	O
double	O
.	O

It	O
does	O
so	O
for	O
each	O
and	O
every	O
app	O
on	O
the	O
device	O
as	O
long	O
as	O
the	O
package	O
names	O
are	O
on	O
its	O
prey	O
list	O
.	O

Over	O
time	O
,	O
this	O
campaign	O
will	O
also	O
infect	O
the	O
same	O
device	O
,	O
repeatedly	O
,	O
with	O
the	O
latest	O
malicious	O
patches	O
.	O

This	O
lead	O
us	O
to	O
estimate	O
there	O
to	O
be	O
over	O
2.8	O
billion	O
infections	O
in	O
total	O
,	O
on	O
around	O
25	O
Million	O
unique	O
devices	O
,	O
meaning	O
that	O
on	O
average	O
,	O
each	O
victim	O
would	O
have	O
suffered	O
roughly	O
112	O
swaps	O
of	O
innocent	O
applications	O
.	O

As	O
an	O
initial	O
attack	O
vector	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
abuses	O
the	O
9Apps	B-System
market	O
–	O
with	O
over	O
360	O
different	O
dropper	O
variants	O
.	O

To	O
maximize	O
profit	O
,	O
variants	O
with	O
“	O
MinSDK	O
”	O
or	O
“	O
OTA	O
”	O
SDK	O
are	O
present	O
to	O
further	O
infect	O
victims	O
with	O
other	O
adware	O
families	O
.	O

The	O
majority	O
of	O
droppers	O
in	O
9Apps	B-System
are	O
games	O
,	O
while	O
the	O
rest	O
fall	O
into	O
categories	O
of	O
adult	O
entertainment	O
,	O
media	O
player	O
,	O
photo	O
utilities	O
,	O
and	O
system	O
utilities	O
.	O

Figure	O
20	O
:	O
dropper	O
app	O
category	O
distribution	O
Among	O
the	O
vast	O
number	O
of	O
variants	O
,	O
the	O
top	O
5	O
most	O
infectious	O
droppers	O
alone	O
have	O
been	O
downloaded	O
more	O
than	O
7.8	O
million	O
times	O
of	O
the	O
infection	O
operations	O
against	O
innocent	O
applications	O
:	O
Figure	O
21	O
:	O
Top	O
5	O
most	O
infectious	O
droppers	O
The	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
is	O
primarily	O
targeted	O
at	O
Indian	O
users	O
,	O
who	O
represent	O
59	O
%	O
of	O
the	O
impacted	O
population	O
.	O

Unlike	O
previously	O
seen	O
non-GP	O
(	O
Google	B-System
Play	I-System
)	O
centric	O
malware	O
campaigns	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
has	O
a	O
significant	O
impact	O
upon	O
not	O
only	O
developing	O
countries	O
but	O
also	O
some	O
developed	O
countries	O
where	O
GP	O
is	O
readily	O
available	O
.	O

For	O
example	O
,	O
the	O
US	O
(	O
with	O
around	O
303k	O
infections	O
)	O
,	O
Saudi	O
Arabia	O
(	O
245k	O
)	O
,	O
Australia	O
(	O
141k	O
)	O
and	O
the	O
UK	O
(	O
137k	O
)	O
.	O

Figure	O
22	O
:	O
world	O
infection	O
heat	O
map	O
Considering	O
that	O
India	O
is	O
by	O
far	O
the	O
most	O
infected	O
county	O
by	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
,	O
overall	O
compromised	O
device	O
brand	O
distribution	O
is	O
heavily	O
influenced	O
by	O
brand	O
popularity	O
among	O
Indian	O
Android	B-System
users	O
:	O
Figure	O
23	O
:	O
infected	O
brand	O
distribution	O
While	O
most	O
infections	O
occurred	O
on	O
devices	O
running	O
Android	B-System
5	I-System
and	I-System
6	I-System
,	O
we	O
also	O
see	O
a	O
considerable	O
number	O
of	O
successful	O
attacks	O
against	O
newer	O
Android	B-System
versions	O
.	O

It	O
is	O
a	O
worrying	O
observation	O
.	O

AOSP	O
patched	O
the	O
Janus	B-Vulnerability
vulnerability	O
since	O
version	O
7	O
by	O
introducing	O
APK	O
Signature	O
Scheme	O
V2	O
.	O

However	O
,	O
in	O
order	O
to	O
block	O
Janus	B-Vulnerability
abuse	O
,	O
app	O
developers	O
need	O
to	O
sign	O
their	O
apps	O
with	O
the	O
new	O
scheme	O
so	O
that	O
Android	B-System
framework	O
security	O
component	O
could	O
conduct	O
integrity	O
checks	O
with	O
enhanced	O
features	O
.	O

Figure	O
25	O
:	O
infected	O
Android	B-System
version	O
distribution	O
To	O
further	O
analyze	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
’	O
s	O
infection	O
landscape	O
,	O
we	O
dived	O
into	O
the	O
top	O
10	O
infected	O
countries	O
:	O
Country	O
Total	O
Devices	O
Total	O
Infection	O
Event	O
Count	O
Avg	O
.	O

App	O
Swap	O
Per	O
Device	O
Avg	O
.	O

Droppers	O
Per	O
Device	O
Avg	O
.	O

Months	O
Device	O
Remained	O
Infected	O
India	O
15,230,123	O
2,017,873,249	O
2.6	O
1.7	O
2.1	O
Bangladesh	O
2,539,913	O
208,026,886	O
2.4	O
1.5	O
2.2	O
Pakistan	O
1,686,216	O
94,296,907	O
2.4	O
1.6	O
2	O
Indonesia	O
572,025	O
67,685,983	O
2	O
1.5	O
2.2	O
Nepal	O
469,274	O
44,961,341	O
2.4	O
1.6	O
2.4	O
US	O
302,852	O
19,327,093	O
1.7	O
1.4	O
1.8	O
Nigeria	O
287,167	O
21,278,498	O
2.4	O
1.3	O
2.3	O
Hungary	O
282,826	O
7,856,064	O
1.7	O
1.3	O
1.7	O
Saudi	O
Arabia	O
245,698	O
18,616,259	O
2.3	O

1.6	O
1.9	O
Myanmar	O
234,338	O
9,729,572	O
1.5	O
1.4	O
1.9	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
Timeline	O
Early	O
signs	O
of	O
activity	O
from	O
the	O
actor	O
behind	O
“	O
Agent	O
Smith	O
”	O
can	O
be	O
traced	O
back	O
to	O
January	O
2016	O
.	O

We	O
classify	O
this	O
40-month	O
period	O
into	O
three	O
main	O
stages	O
.	O

January	O
2016	O
–	O
May	O
2018	O
:	O
In	O
this	O
stage	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
hackers	O
started	O
to	O
try	O
out	O
9Apps	O
as	O
a	O
distribution	O
channel	O
for	O
their	O
adware	O
.	O

During	O
this	O
period	O
,	O
malware	O
samples	O
display	O
some	O
typical	O
adware	O
characteristics	O
such	O
as	O
unnecessary	O
permission	O
requirements	O
and	O
pop-up	O
windows	B-System
.	O

During	O
this	O
time	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
hackers	O
eventually	O
built	O
up	O
a	O
vast	O
number	O
of	O
app	O
presence	O
on	O
9Apps	B-System
,	O
which	O
later	O
would	O
serve	O
as	O
publication	O
channels	O
for	O
evolved	O
droppers	O
.	O

However	O
,	O
samples	O
don	O
’	O
t	O
have	O
key	O
capabilities	O
to	O
infect	O
innocent	O
apps	O
on	O
victim	O
devices	O
yet	O
.	O

May	O
2018	O
to	O
April	O
2019	O
:	O
This	O
is	O
the	O
actual	O
mature	O
stage	O
of	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
.	O

From	O
early	O
2018	O
prior	O
to	O
May	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
hackers	O
started	O
to	O
experiment	O
with	O
Bundle	O
Feng	O
Shui	O
,	O
the	O
key	O
tool	O
which	O
gives	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
malware	O
family	O
capabilities	O
to	O
infect	O
innocent	O
apps	O
on	O
the	O
device	O
.	O

A	O
series	O
of	O
pilot	O
runs	O
were	O
executed	O
.	O

After	O
some	O
major	O
upgrade	O
,	O
by	O
mid-June	O
,	O
the	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
reached	O
its	O
peak	O
.	O

Its	O
dropper	O
family	O
finished	O
integration	O
with	O
Bundle	O
Feng	O
Shui	O
and	O
campaign	O
C	O
&	O
C	O
infrastructure	O
was	O
shifted	O
to	O
AWS	B-System
cloud	O
.	O

The	O
Campaign	O
achieved	O
exponential	O
growth	O
from	O
June	O
to	O
December	O
2018	O
with	O
the	O
infection	O
number	O
staying	O
stable	O
into	O
early	O
2019	O
.	O

Post-April	O
2019	O
:	O
Starting	O
from	O
early	O
2019	O
,	O
the	O
new	O
infection	O
rate	O
of	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
dropped	O
significantly	O
.	O

From	O
early	O
April	O
,	O
hackers	O
started	O
to	O
build	O
a	O
new	O
major	O
update	O
to	O
the	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
under	O
the	O
name	O
“	O
leechsdk	O
”	O
.	O

Figure	O
26	O
:	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
Campaign	O
timeline	O
Greater	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
Campaign	O
Discovery	O
Orchestrating	O
a	O
successful	O
9Apps	B-System
centric	O
malware	O
campaign	O
,	O
the	O
actor	O
behind	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
established	O
solid	O
strategies	O
in	O
malware	O
proliferation	O
and	O
payload	O
delivery	O
.	O

The	O
actor	O
also	O
built	O
solid	O
backend	O
infrastructures	O
which	O
can	O
handle	O
high	O
volume	O
concurrent	O
requests	O
.	O

During	O
our	O
extended	O
threat	O
hunting	O
,	O
we	O
uncovered	O
11	O
apps	O
on	O
the	O
Google	B-System
Play	I-System
store	I-System
that	O
contain	O
a	O
malicious	O
yet	O
dormant	O
SDK	O
related	O
to	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
actor	O
.	O

This	O
discovery	O
indicates	O
the	O
actor	O
’	O
s	O
ambition	O
in	O
expanding	O
operations	O
into	O
Google	B-System
Play	I-System
store	O
with	O
previous	O
success	O
experience	O
from	O
the	O
main	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
.	O

Instead	O
of	O
embedding	O
core	O
malware	O
payload	O
in	O
droppers	O
,	O
the	O
actor	O
switches	O
to	O
a	O
more	O
low-key	O
SDK	O
approach	O
.	O

In	O
the	O
dangerous	O
module	O
lies	O
a	O
kill	O
switch	O
logic	O
which	O
looks	O
for	O
the	O
keyword	O
“	O
infect	O
”	O
.	O

Once	O
the	O
keyword	O
is	O
present	O
,	O
the	O
SDK	O
will	O
switch	O
from	O
innocent	O
ads	O
server	O
to	O
malicious	O
payload	O
delivery	O
ones	O
.	O

Hence	O
,	O
we	O
name	O
this	O
new	O
spin-off	O
campaign	O
as	O
Jaguar	O
Kill	O
Switch	O
.	O

The	O
below	O
code	O
snippet	O
is	O
currently	O
isolated	O
and	O
dormant	O
.	O

In	O
the	O
future	O
,	O
it	O
will	O
be	O
invoked	O
by	O
malicious	O
SDK	O
during	O
banner	O
ads	O
display	O
.	O

Figure	O
26	O
:	O
the	O
kill	O
switch	O
code	O
snippet	O
Evidence	O
implies	O
that	O
the	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
actor	O
is	O
currently	O
laying	O
the	O
groundwork	O
,	O
increasing	O
its	O
Google	B-System
Play	I-System
penetration	O
rate	O
and	O
waiting	O
for	O
the	O
right	O
timing	O
to	O
kick	O
off	O
attacks	O
.	O

By	O
the	O
time	O
of	O
this	O
publication	O
,	O
two	O
Jaguar	O
Kill	O
Switch	O
infected	O
app	O
has	O
reached	O
10	O
million	O
downloads	O
while	O
others	O
are	O
still	O
in	O
their	O
early	O
stages	O
.	O

Check	B-Organization
Point	I-Organization
Research	O
reported	O
these	O
dangerous	O
apps	O
to	O
Google	B-Organization
upon	O
discovery	O
.	O

Currently	O
,	O
all	O
bespoke	O
apps	O
have	O
been	O
taken	O
down	O
from	O
the	O
Google	B-System
Play	I-System
store	O
.	O

Figure	O
28	O
:	O
Jaguar	O
Kill	O
Switch	O
infected	O
GP	O
apps	O
Peek	O
Into	O
the	O
Actor	O
Based	O
on	O
all	O
of	O
the	O
above	O
,	O
we	O
connected	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
to	O
a	O
Chinese	O
internet	O
company	O
located	O
in	O
Guangzhou	O
whose	O
front	O
end	O
legitimate	O
business	O
is	O
to	O
help	O
Chinese	O
Android	B-System
developers	O
publish	O
and	O
promote	O
their	O
apps	O
on	O
overseas	O
platforms	O
.	O

Various	O
recruitment	O
posts	O
on	O
Chinese	O
job	O
sites	O
and	O
Chinese	B-System
National	I-System
Enterprise	I-System
Credit	I-System
Information	I-System
Public	I-System
System	I-System
(	I-System
NECIPS	I-System
)	I-System
data	O
led	O
us	O
one	O
step	O
further	O
,	O
linking	O
the	O
actor	O
to	O
its	O
legal	O
entity	O
name	O
.	O

Interestingly	O
,	O
we	O
uncovered	O
several	O
expired	O
job	O
posting	O
of	O
Android	B-System
reverse	O
engineer	O
from	O
the	O
actor	O
’	O
s	O
front	O
business	O
published	O
in	O
2018	O
and	O
2019	O
.	O

It	O
seems	O
that	O
the	O
people	O
who	O
filled	O
these	O
roles	O
are	O
key	O
to	O
“	O
Agent	B-Malware
Smith	I-Malware
’	O
s	O
success	O
,	O
yet	O
not	O
quite	O
necessary	O
for	O
actor	O
’	O
s	O
legitimate	O
side	O
of	O
business	O
.	O

With	O
a	O
better	O
understanding	O
of	O
the	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
actor	O
than	O
we	O
had	O
in	O
the	O
initial	O
phase	O
of	O
campaign	O
hunting	O
,	O
we	O
examined	O
the	O
list	O
of	O
target	O
innocent	O
apps	O
once	O
again	O
and	O
discovered	O
the	O
actor	O
’	O
s	O
unusual	O
practices	O
in	O
choosing	O
targets	O
.	O

It	O
seems	O
,	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
prey	O
list	O
does	O
not	O
only	O
have	O
popular	O
yet	O
Janus	B-Vulnerability
vulnerable	O
apps	O
to	O
ensure	O
high	O
proliferation	O
,	O
but	O
also	O
contain	O
competitor	O
apps	O
of	O
actor	O
’	O
s	O
legitimate	O
business	O
arm	O
to	O
suppress	O
competition	O
.	O

Conclusion	O
Although	O
the	O
actor	O
behind	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
decided	O
to	O
make	O
their	O
illegally	O
acquired	O
profit	O
by	O
exploiting	O
the	O
use	O
of	O
ads	O
,	O
another	O
actor	O
could	O
easily	O
take	O
a	O
more	O
intrusive	O
and	O
harmful	O
route	O
.	O

With	O
the	O
ability	O
to	O
hide	O
its	O
icon	O
from	O
the	O
launcher	O
and	O
hijack	O
popular	O
existing	O
apps	O
on	O
a	O
device	O
,	O
there	O
are	O
endless	O
possibilities	O
to	O
harm	O
a	O
user	O
’	O
s	O
digital	O
even	O
physical	O
security	O
.	O

Today	O
this	O
malware	O
shows	O
unwanted	O
ads	O
,	O
tomorrow	O
it	O
could	O
steal	O
sensitive	O
information	O
;	O
from	O
private	O
messages	O
to	O
banking	O
credentials	O
and	O
much	O
more	O
.	O

The	O
“	O
Agent	B-Malware
Smith	I-Malware
”	O
campaign	O
serves	O
as	O
a	O
sharp	O
reminder	O
that	O
effort	O
from	O
system	O
developers	O
alone	O
is	O
not	O
enough	O
to	O
build	O
a	O
secure	O
Android	B-System
eco-system	O
.	O

It	O
requires	O
attention	O
and	O
action	O
from	O
system	O
developers	O
,	O
device	O
manufacturers	O
,	O
app	O
developers	O
,	O
and	O
users	O
,	O
so	O
that	O
vulnerability	O
fixes	O
are	O
patched	O
,	O
distributed	O
,	O
adopted	O
and	O
installed	O
in	O
time	O
.	O

It	O
is	O
also	O
another	O
example	O
for	O
why	O
organizations	O
and	O
consumers	O
alike	O
should	O
have	O
an	O
advanced	O
mobile	O
threat	O
prevention	O
solution	O
installed	O
on	O
the	O
device	O
to	O
protect	O
themselves	O
against	O
the	O
possibility	O
of	O
unknowingly	O
installing	O
malicious	O
apps	O
,	O
even	O
from	O
trusted	O
app	O
stores	O
.	O

Dvmap	B-Malware
:	O
the	O
first	O
Android	B-System
malware	O
with	O
code	O
injection	O
08	O
JUN	O
2017	O
In	O
April	O
2017	O
we	O
started	O
observing	O
new	O
rooting	O
malware	O
being	O
distributed	O
through	O
the	O
Google	B-System
Play	I-System
Store	I-System
.	O

Unlike	O
other	O
rooting	O
malware	O
,	O
this	O
Trojan	O
not	O
only	O
installs	O
its	O
modules	O
into	O
the	O
system	O
,	O
it	O
also	O
injects	O
malicious	O
code	O
into	O
the	O
system	O
runtime	O
libraries	O
.	O

Kaspersky	B-Organization
Lab	I-Organization
products	O
detect	O
it	O
as	O
Trojan.AndroidOS.Dvmap.a	B-Indicator
.	O

The	O
distribution	O
of	O
rooting	O
malware	O
through	O
Google	B-System
Play	I-System
is	O
not	O
a	O
new	O
thing	O
.	O

For	O
example	O
,	O
the	O
Ztorg	B-Malware
Trojan	I-Malware
has	O
been	O
uploaded	O
to	O
Google	B-System
Play	I-System
almost	O
100	O
times	O
since	O
September	O
2016	O
.	O

But	O
Dvmap	B-Malware
is	O
very	O
special	O
rooting	O
malware	O
.	O

It	O
uses	O
a	O
variety	O
of	O
new	O
techniques	O
,	O
but	O
the	O
most	O
interesting	O
thing	O
is	O
that	O
it	O
injects	O
malicious	O
code	O
into	O
the	O
system	O
libraries	O
–	O
libdmv.so	B-Indicator
or	O
libandroid_runtime.so	B-Indicator
.	O

This	O
makes	O
Dvmap	B-Malware
the	O
first	O
Android	B-System
malware	O
that	O
injects	O
malicious	O
code	O
into	O
the	O
system	O
libraries	O
in	O
runtime	O
,	O
and	O
it	O
has	O
been	O
downloaded	O
from	O
the	O
Google	B-System
Play	I-System
Store	I-System
more	O
than	O
50,000	O
times	O
.	O

Kaspersky	B-Organization
Lab	I-Organization
reported	O
the	O
Trojan	O
to	O
Google	B-Organization
,	O
and	O
it	O
has	O
now	O
been	O
removed	O
from	O
the	O
store	O
.	O

To	O
bypass	O
Google	B-System
Play	I-System
Store	I-System
security	O
checks	O
,	O
the	O
malware	O
creators	O
used	O
a	O
very	O
interesting	O
method	O
:	O
they	O
uploaded	O
a	O
clean	O
app	O
to	O
the	O
store	O
at	O
the	O
end	O
of	O
March	O
,	O
2017	O
,	O
and	O
would	O
then	O
update	O
it	O
with	O
a	O
malicious	O
version	O
for	O
short	O
period	O
of	O
time	O
.	O

Usually	O
they	O
would	O
upload	O
a	O
clean	O
version	O
back	O
on	O
Google	B-System
Play	I-System
the	O
very	O
same	O
day	O
.	O

They	O
did	O
this	O
at	O
least	O
5	O
times	O
between	O
18	O
April	O
and	O
15	O
May	O
.	O

All	O
the	O
malicious	O
Dvmap	B-Malware
apps	O
had	O
the	O
same	O
functionality	O
.	O

They	O
decrypt	O
several	O
archive	O
files	O
from	O
the	O
assets	O
folder	O
of	O
the	O
installation	O
package	O
,	O
and	O
launch	O
an	O
executable	O
file	O
from	O
them	O
with	O
the	O
name	O
“	O
start.	O
”	O
The	O
interesting	O
thing	O
is	O
that	O
the	O
Trojan	O
supports	O
even	O
the	O
64-bit	O
version	O
of	O
Android	B-System
,	O
which	O
is	O
very	O
rare	O
.	O

All	O
encrypted	O
archives	O
can	O
be	O
divided	O
into	O
two	O
groups	O
:	O
the	O
first	O
comprises	O
Game321.res	B-Indicator
,	O
Game322.res	B-Indicator
,	O
Game323.res	B-Indicator
and	O
Game642.res	B-Indicator
–	O
and	O
these	O
are	O
used	O
in	O
the	O
initial	O
phase	O
of	O
infection	O
,	O
while	O
the	O
second	O
group	O
:	O
Game324.res	B-Indicator
and	O
Game644.res	B-Indicator
,	O
are	O
used	O
in	O
the	O
main	O
phase	O
.	O

Initial	O
phase	O
During	O
this	O
phase	O
,	O
the	O
Trojan	O
tries	O
to	O
gain	O
root	O
rights	O
on	O
the	O
device	O
and	O
to	O
install	O
some	O
modules	O
.	O

All	O
archives	O
from	O
this	O
phase	O
contain	O
the	O
same	O
files	O
except	O
for	O
one	O
called	O
“	O
common	O
”	O
.	O

This	O
is	O
a	O
local	O
root	O
exploit	O
pack	O
,	O
and	O
the	O
Trojan	O
uses	O
4	O
different	O
exploit	O
pack	O
files	O
,	O
3	O
for	O
32-bit	O
systems	O
and	O
1	O
for	O
64-bit-systems	O
.	O

If	O
these	O
files	O
successfully	O
gain	O
root	O
rights	O
,	O
the	O
Trojan	O
will	O
install	O
several	O
tools	O
into	O
the	O
system	O
.	O

It	O
will	O
also	O
install	O
the	O
malicious	O
app	O
“	O
com.qualcmm.timeservices.	B-Indicator
”	O
These	O
archives	O
contain	O
the	O
file	O
“	O
.root.sh	B-Indicator
”	O
which	O
has	O
some	O
comments	O
in	O
Chinese	O
:	O
Main	O
phase	O
In	O
this	O
phase	O
,	O
the	O
Trojan	O
launches	O
the	O
“	O
start	O
”	O
file	O
from	O
Game324.res	B-Indicator
or	O
Game644.res	B-Indicator
.	O

It	O
will	O
check	O
the	O
version	O
of	O
Android	B-System
installed	O
and	O
decide	O
which	O
library	O
should	O
be	O
patched	O
.	O

For	O
Android	B-System
4.4.4	I-System
and	O
older	O
,	O
the	O
Trojan	O
will	O
patch	O
method	O
_Z30dvmHeapSourceStartupBeforeForkv	O
from	O
libdvm.so	B-Indicator
,	O
and	O
for	O
Android	B-System
5	O
and	O
newer	O
it	O
will	O
patch	O
method	O
nativeForkAndSpecialize	O
from	O
libandroid_runtime.so	B-Indicator
.	O

Both	O
of	O
these	O
libraries	O
are	O
runtime	O
libraries	O
related	O
to	O
Dalvik	B-System
and	O
ART	B-System
runtime	O
environments	O
.	O

Before	O
patching	O
,	O
the	O
Trojan	O
will	O
backup	O
the	O
original	O
library	O
with	O
a	O
name	O
bak_	O
{	O
original	O
name	O
}	O
.	O

During	O
patching	O
,	O
the	O
Trojan	O
will	O
overwrite	O
the	O
existing	O
code	O
with	O
malicious	O
code	O
so	O
that	O
all	O
it	O
can	O
do	O
is	O
execute	O
/system/bin/ip	B-Indicator
.	O

This	O
could	O
be	O
very	O
dangerous	O
and	O
cause	O
some	O
devices	O
to	O
crash	O
following	O
the	O
overwrite	O
.	O

Then	O
the	O
Trojan	O
will	O
put	O
the	O
patched	O
library	O
back	O
into	O
the	O
system	O
directory	O
.	O

After	O
that	O
,	O
the	O
Trojan	O
will	O
replace	O
the	O
original	O
/system/bin/ip	B-Indicator
with	O
a	O
malicious	O
one	O
from	O
the	O
archive	O
(	O
Game324.res	B-Indicator
or	O
Game644.res	B-Indicator
)	O
.	O

In	O
doing	O
so	O
,	O
the	O
Trojan	O
can	O
be	O
sure	O
that	O
its	O
malicious	O
module	O
will	O
be	O
executed	O
with	O
system	O
rights	O
.	O

But	O
the	O
malicious	O
ip	O
file	O
does	O
not	O
contain	O
any	O
methods	O
from	O
the	O
original	O
ip	O
file	O
.	O

This	O
means	O
that	O
all	O
apps	O
that	O
were	O
using	O
this	O
file	O
will	O
lose	O
some	O
functionality	O
or	O
even	O
start	O
crashing	O
.	O

Malicious	O
module	O
“	O
ip	O
”	O
This	O
file	O
will	O
be	O
executed	O
by	O
the	O
patched	O
system	O
library	O
.	O

It	O
can	O
turn	O
off	O
“	O
VerifyApps	O
”	O
and	O
enable	O
the	O
installation	O
of	O
apps	O
from	O
3rd	O
party	O
stores	O
by	O
changing	O
system	O
settings	O
.	O

Furthermore	O
,	O
it	O
can	O
grant	O
the	O
“	O
com.qualcmm.timeservices	B-Indicator
”	O
app	O
Device	O
Administrator	O
rights	O
without	O
any	O
interaction	O
with	O
the	O
user	O
,	O
just	O
by	O
running	O
commands	O
.	O

It	O
is	O
a	O
very	O
unusual	O
way	O
to	O
get	O
Device	O
Administrator	O
rights	O
.	O

Malicious	O
app	O
com.qualcmm.timeservices	B-Indicator
As	O
I	O
mentioned	O
before	O
,	O
in	O
the	O
“	O
initial	O
phase	O
”	O
,	O
the	O
Trojan	O
will	O
install	O
the	O
“	O
com.qualcmm.timeservices	B-Indicator
”	O
app	O
.	O

Its	O
main	O
purpose	O
is	O
to	O
download	O
archives	O
and	O
execute	O
the	O
“	O
start	O
”	O
binary	O
from	O
them	O
.	O

During	O
the	O
investigation	O
,	O
this	O
app	O
was	O
able	O
to	O
successfully	O
connect	O
to	O
the	O
command	O
and	O
control	O
server	O
,	O
but	O
it	O
received	O
no	O
commands	O
.	O

So	O
I	O
don	O
’	O
t	O
know	O
what	O
kind	O
of	O
files	O
will	O
be	O
executed	O
,	O
but	O
they	O
could	O
be	O
malicious	O
or	O
advertising	O
files	O
.	O

Conclusions	O
This	O
Trojan	O
was	O
distributed	O
through	O
the	O
Google	B-System
Play	I-System
Store	I-System
and	O
uses	O
a	O
number	O
of	O
very	O
dangerous	O
techniques	O
,	O
including	O
patching	O
system	O
libraries	O
.	O

It	O
installs	O
malicious	O
modules	O
with	O
different	O
functionality	O
into	O
the	O
system	O
.	O

It	O
looks	O
like	O
its	O
main	O
purpose	O
is	O
to	O
get	O
into	O
the	O
system	O
and	O
execute	O
downloaded	O
files	O
with	O
root	O
rights	O
.	O

But	O
I	O
never	O
received	O
such	O
files	O
from	O
their	O
command	O
and	O
control	O
server	O
.	O

These	O
malicious	O
modules	O
report	O
to	O
the	O
attackers	O
about	O
every	O
step	O
they	O
are	O
going	O
to	O
make	O
.	O

So	O
I	O
think	O
that	O
the	O
authors	O
are	O
still	O
testing	O
this	O
malware	O
,	O
because	O
they	O
use	O
some	O
techniques	O
which	O
can	O
break	O
the	O
infected	O
devices	O
.	O

But	O
they	O
already	O
have	O
a	O
lot	O
of	O
infected	O
users	O
on	O
whom	O
to	O
test	O
their	O
methods	O
.	O

I	O
hope	O
that	O
by	O
uncovering	O
this	O
malware	O
at	O
such	O
an	O
early	O
stage	O
,	O
we	O
will	O
be	O
able	O
to	O
prevent	O
a	O
massive	O
and	O
dangerous	O
attack	O
when	O
the	O
attackers	O
are	O
ready	O
to	O
actively	O
use	O
their	O
methods	O
.	O

MD5	O
43680D1914F28E14C90436E1D42984E2	B-Indicator
20D4B9EB9377C499917C4D69BF4CCEBE	B-Indicator
First	O
widely	O
distributed	O
Android	B-System
bootkit	O
Malware	O
infects	O
more	O
than	O
350,000	O
Devices	O
January	O
29	O
,	O
2014	O
In	O
the	O
last	O
quarter	O
of	O
2013	O
,	O
sale	O
of	O
a	O
Smartphone	O
with	O
ANDROID	B-System
operating	O
system	O
has	O
increased	O
and	O
every	O
second	O
person	O
you	O
see	O
is	O
a	O
DROID	B-System
user	O
.	O

A	O
Russian	O
security	O
firm	O
'Doctor	O
Web	B-Organization
'	O
identified	O
the	O
first	O
mass	O
distributed	O
Android	B-System
bootkit	O
malware	O
called	O
'Android.Oldboot	O
'	O
,	O
a	O
piece	O
of	O
malware	O
that	O
's	O
designed	O
to	O
re-infect	O
devices	O
after	O
reboot	O
,	O
even	O
if	O
you	O
delete	O
all	O
working	O
components	O
of	O
it	O
.	O

The	O
bootkit	O
Android.Oldboot	B-Malware
has	O
infected	O
more	O
than	O
350,000	O
android	B-System
users	O
in	O
China	O
,	O
Spain	O
,	O
Italy	O
,	O
Germany	O
,	O
Russia	O
,	O
Brazil	O
,	O
the	O
USA	O
and	O
some	O
Southeast	O
Asian	O
countries	O
.	O

China	O
seems	O
to	O
a	O
mass	O
victim	O
of	O
this	O
kind	O
of	O
malware	O
having	O
a	O
92	O
%	O
share	O
.	O

A	O
Bootkit	O
is	O
a	O
rootkit	O
malware	O
variant	O
which	O
infects	O
the	O
device	O
at	O
start-up	O
and	O
may	O
encrypt	O
disk	O
or	O
steal	O
data	O
,	O
remove	O
the	O
application	O
,	O
open	O
connection	O
for	O
Command	O
and	O
controller	O
.	O

A	O
very	O
unique	O
technique	O
is	O
being	O
used	O
to	O
inject	O
this	O
Trojan	O
into	O
an	O
Android	B-System
system	O
where	O
an	O
attacker	O
places	O
a	O
component	O
of	O
it	O
into	O
the	O
boot	O
partition	O
of	O
the	O
file	O
system	O
and	O
modify	O
the	O
'init	O
'	O
script	O
(	O
initialize	O
the	O
operating	O
system	O
)	O
to	O
re-load	O
the	O
malware	O
as	O
you	O
switch	O
on	O
your	O
android	B-System
.	O

When	O
you	O
start	O
your	O
device	O
,	O
this	O
script	O
loads	O
the	O
Trojan	O
'imei_chk	O
'	O
(	O
detects	O
it	O
as	O
Android.Oldboot.1	B-Indicator
)	O
which	O
extract	O
two	O
files	O
libgooglekernel.so	B-Indicator
(	O
Android.Oldboot.2	B-Indicator
)	O
and	O
GoogleKernel.apk	B-Indicator
(	O
Android.Oldboot.1.origin	B-Indicator
)	O
,	O
copy	O
them	O
respectively	O
in	O
/system/lib	B-Indicator
and	I-Indicator
/system/app	I-Indicator
.	O

Android.Oldboot	B-Malware
acts	O
as	O
a	O
system	O
service	O
and	O
connects	O
to	O
the	O
command-and-controller	O
server	O
using	O
libgooglekernel.so	B-Indicator
library	O
and	O
receives	O
commands	O
to	O
download	O
,	O
remove	O
installed	O
apps	O
,	O
and	O
install	O
malicious	O
apps	O
.	O

Since	O
it	O
becomes	O
a	O
part	O
of	O
the	O
boot	O
partition	O
,	O
formatting	O
the	O
device	O
will	O
not	O
solve	O
the	O
problem	O
.	O

The	O
researchers	O
believe	O
that	O
the	O
devices	O
somehow	O
had	O
the	O
malware	O
pre-loaded	O
at	O
the	O
time	O
of	O
shipping	O
from	O
the	O
manufacturer	O
,	O
or	O
was	O
likely	O
distributed	O
inside	O
modified	O
Android	B-System
firmware	O
.	O

So	O
,	O
users	O
should	O
beware	O
of	O
certain	O
modified	O
Android	B-System
firmware	O
.	O

Two	O
weeks	O
ago	O
,	O
Some	O
Chinese	O
Security	O
Researchers	O
have	O
also	O
detected	O
a	O
bootkit	O
called	O
'Oldboot	O
'	O
,	O
possibly	O
the	O
same	O
malware	O
or	O
another	O
variant	O
of	O
it	O
.	O

"	O
Due	O
to	O
the	O
special	O
RAM	O
disk	O
feature	O
of	O
Android	B-System
devices	O
'	O
boot	O
partition	O
,	O
all	O
current	O
mobile	O
antivirus	O
products	O
in	O
the	O
world	O
ca	O
n't	O
completely	O
remove	O
this	O
Trojan	O
or	O
effectively	O
repair	O
the	O
system	O
.	O

''	O
"	O
According	O
to	O
our	O
statistics	O
,	O
as	O
of	O
today	O
,	O
there	O
're	O
more	O
than	O
500	O
,	O
000	O
Android	B-System
devices	O
infected	O
by	O
this	O
bootkit	O
in	O
China	O
in	O
last	O
six	O
months	O
.	O

The	O
Android	B-System
malware	O
Android.Oldboot	B-Malware
is	O
almost	O
impossible	O
to	O
remove	O
,	O
not	O
even	O
with	O
formatting	O
your	O
device	O
.	O

But	O
if	O
your	O
device	O
is	O
not	O
from	O
a	O
Chinese	O
manufacturer	O
,	O
then	O
chances	O
that	O
you	O
are	O
a	O
victim	O
of	O
it	O
,	O
are	O
very	O
less	O
.	O

This	O
bootkit	O
is	O
not	O
the	O
first	O
of	O
this	O
kind	O
.	O

Two	O
years	O
back	O
,	O
in	O
the	O
month	O
of	O
March	O
we	O
reported	O
,	O
NQ	B-Organization
Mobile	I-Organization
Security	I-Organization
Research	O
Center	O
uncovered	O
the	O
world	O
's	O
first	O
Android	B-System
bootkit	O
malware	O
called	O
'DKFBootKit	O
'	O
,	O
that	O
replaces	O
certain	O
boot	O
processes	O
and	O
can	O
begin	O
running	O
even	O
before	O
the	O
system	O
is	O
completely	O
booted	O
up	O
.	O

But	O
Android.Oldboot	B-Malware
malware	O
is	O
a	O
bit	O
more	O
dangerous	O
because	O
even	O
if	O
you	O
remove	O
all	O
working	O
components	O
of	O
it	O
from	O
your	O
android	B-System
successfully	O
,	O
the	O
component	O
imei_chk	B-Indicator
will	O
persist	O
in	O
a	O
protected	O
boot	O
memory	O
area	O
and	O
hence	O
will	O
reinstall	O
itself	O
on	O
next	O
boot	O
and	O
continuously	O
infect	O
the	O
Smartphone	O
.	O

Users	O
are	O
recommended	O
to	O
install	O
apps	O
from	O
authorized	O
stores	O
such	O
as	O
Google	B-System
Play	I-System
,	O
disable	O
installation	O
of	O
apps	O
from	O
'Unknown	O
Sources	O
'	O
and	O
for	O
a	O
better	O
security	O
install	O
a	O
reputed	O
security	O
application	O
.	O

You	O
can	O
also	O
try	O
to	O
re-flash	O
your	O
device	O
with	O
its	O
original	O
ROM	O
.	O

After	O
flashing	O
,	O
the	O
bootkit	O
will	O
be	O
removed	O
.	O

FrozenCell	B-Malware
:	O
Multi-Platform	O
Surveillance	O
Campaign	O
Against	O
Palestinians	O
October	O
5	O
,	O
2017	O
FrozenCell	B-Malware
has	O
been	O
seen	O
masquerading	O
as	O
various	O
well	O
known	O
social	O
media	O
and	O
chat	O
applications	O
as	O
well	O
as	O
an	O
app	O
likely	O
only	O
used	O
by	O
Palestinian	O
or	O
Jordanian	O
students	O
sitting	O
their	O
2016	O
general	O
exams	O
.	O

Lookout	B-Organization
researchers	O
have	O
discovered	O
a	O
new	O
mobile	O
surveillanceware	O
family	O
,	O
FrozenCell	B-Malware
.	O

The	O
threat	O
is	O
likely	O
targeting	O
employees	O
of	O
various	O
Palestinian	O
government	O
agencies	O
,	O
security	O
services	O
,	O
Palestinian	O
students	O
,	O
and	O
those	O
affiliated	O
with	O
the	O
Fatah	B-Organization
political	O
party	O
.	O

FrozenCell	B-Malware
is	O
the	O
mobile	O
component	O
of	O
a	O
multi-platform	O
attack	O
we	O
've	O
seen	O
a	O
threat	O
actor	O
known	O
as	O
"	O
Two-tailed	B-Malware
Scorpion/APT-C-23	I-Malware
,	O
''	O
use	O
to	O
spy	O
on	O
victims	O
through	O
compromised	O
mobile	O
devices	O
and	O
desktops	O
.	O

The	O
desktop	O
components	O
of	O
this	O
attack	O
,	O
previously	O
discovered	O
by	O
Palo	B-Organization
Alto	I-Organization
Network	I-Organization
,	O
are	O
known	O
as	O
KasperAgent	B-Malware
and	O
Micropsia	B-Malware
.	O

We	O
discovered	O
561MB	O
of	O
exfiltrated	O
data	O
from	O
24	O
compromised	O
Android	B-System
devices	O
while	O
investigating	O
this	O
threat	O
.	O

More	O
data	O
is	O
appearing	O
daily	O
,	O
leading	O
us	O
to	O
believe	O
the	O
actors	O
are	O
still	O
highly	O
active	O
.	O

We	O
are	O
continuing	O
to	O
watch	O
it	O
closely	O
.	O

This	O
threat	O
is	O
another	O
proof	O
point	O
that	O
attackers	O
are	O
clearly	O
incorporating	O
the	O
mobile	O
device	O
into	O
their	O
surveillance	O
campaigns	O
as	O
a	O
primary	O
attack	O
vector	O
.	O

Government	O
agencies	O
and	O
enterprises	O
should	O
look	O
at	O
this	O
threat	O
as	O
an	O
example	O
of	O
the	O
kind	O
of	O
spying	O
that	O
is	O
now	O
possible	O
given	O
how	O
ubiquitous	O
mobile	O
devices	O
are	O
in	O
the	O
workplace	O
.	O

Attackers	O
are	O
keenly	O
aware	O
of	O
the	O
information	O
they	O
can	O
derive	O
from	O
these	O
devices	O
and	O
are	O
using	O
multi-stage	O
(	O
phishing	O
+	O
an	O
executable	O
)	O
,	O
multi-platform	O
(	O
Android	B-System
+	O
desktop	O
)	O
attacks	O
to	O
accomplish	O
their	O
spying	O
.	O

All	O
Lookout	B-Organization
customers	O
are	O
protected	O
from	O
this	O
threat	O
.	O

What	O
it	O
does	O
FrozenCell	B-Malware
masquerades	O
as	O
fake	O
updates	O
to	O
chat	O
applications	O
like	O
Facebook	B-System
,	O
WhatsApp	B-System
,	O
Messenger	B-System
,	O
LINE	B-System
,	O
and	O
LoveChat	B-System
.	O

We	O
also	O
detected	O
it	O
in	O
apps	O
targeted	O
toward	O
specific	O
Middle	O
Eastern	O
demographics	O
.	O

For	O
example	O
,	O
the	O
actors	O
behind	O
FrozenCell	B-Malware
used	O
a	O
spoofed	O
app	O
called	O
Tawjihi	B-Indicator
2016	I-Indicator
,	O
which	O
Jordanian	O
or	O
Palestinian	O
students	O
would	O
ordinarily	O
use	O
during	O
their	O
general	O
secondary	O
examination	O
.	O

Once	O
installed	O
on	O
a	O
device	O
FrozenCell	B-Malware
is	O
capable	O
of	O
:	O
Recording	O
calls	O
Retrieving	O
generic	O
phone	O
metadata	O
(	O
e.g.	O
,	O
cell	O
location	O
,	O
mobile	O
country	O
code	O
,	O
mobile	O
network	O
code	O
)	O
Geolocating	O
a	O
device	O
Extracting	O
SMS	O
messages	O
Retrieving	O
a	O
victim	O
's	O
accounts	O
Exfiltrating	O
images	O
Downloading	O
and	O
installing	O
additional	O
applications	O
Searching	O
for	O
and	O
exfiltrating	O
pdf	O
,	O
doc	O
,	O
docx	O
,	O
ppt	O
,	O
pptx	O
,	O
xls	O
,	O
and	O
xlsx	O
file	O
types	O
Retrieving	O
contacts	O
The	O
graph	O
below	O
represents	O
a	O
split	O
of	O
the	O
types	O
of	O
data	O

from	O
only	O
one	O
misconfigured	O
command	O
and	O
control	O
server	O
(	O
out	O
of	O
over	O
37	O
servers	O
)	O
.	O

This	O
is	O
only	O
a	O
small	O
picture	O
of	O
the	O
threat	O
actor	O
's	O
operations	O
.	O

Split	O
of	O
exfiltrated	O
data	O
Some	O
noteworthy	O
files	O
identified	O
in	O
content	O
taken	O
from	O
compromised	O
devices	O
include	O
passport	O
photos	O
,	O
audio	O
recordings	O
of	O
calls	O
,	O
other	O
images	O
,	O
and	O
a	O
PDF	O
document	O
with	O
data	O
on	O
484	O
individuals	O
.	O

The	O
PDF	O
lists	O
dates	O
of	O
birth	O
,	O
gender	O
,	O
passport	O
numbers	O
,	O
and	O
names	O
.	O

Potential	O
targets	O
The	O
actors	O
behind	O
FrozenCell	B-Malware
used	O
an	O
online	O
service	O
that	O
geolocates	O
mobile	O
devices	O
based	O
on	O
nearby	O
cell	O
towers	O
to	O
track	O
targets	O
.	O

This	O
data	O
shows	O
a	O
distinct	O
concentration	O
of	O
infected	O
devices	O
beaconing	O
from	O
Gaza	O
,	O
Palestine	O
.	O

Map	O
of	O
potential	O
targets	O
Early	O
samples	O
of	O
FrozenCell	B-Malware
used	O
an	O
online	O
service	O
for	O
storing	O
geolocation	O
information	O
of	O
infected	O
devices	O
.	O

Analysis	O
of	O
this	O
telemetry	O
shows	O
infected	O
devices	O
are	O
completely	O
based	O
in	O
Gaza	O
,	O
Palestine	O
.	O

It	O
has	O
not	O
been	O
confirmed	O
whether	O
these	O
are	O
from	O
test	O
devices	O
or	O
the	O
devices	O
of	O
victims	O
.	O

We	O
were	O
also	O
able	O
to	O
link	O
the	O
FrozenCell	B-Malware
's	O
Android	B-System
infrastructure	O
to	O
numerous	O
desktop	O
samples	O
that	O
are	O
part	O
of	O
the	O
larger	O
multi-platform	O
attack	O
.	O

It	O
appears	O
the	O
attackers	O
sent	O
malicious	O
executables	O
though	O
phishing	O
campaigns	O
impersonating	O
individuals	O
associated	O
with	O
the	O
Palestinian	B-Organization
Security	I-Organization
Services	I-Organization
,	O
the	O
General	B-Organization
Directorate	I-Organization
of	I-Organization
Civil	I-Organization
Defence	I-Organization
-	O
Ministry	B-Organization
of	I-Organization
the	I-Organization
Interior	I-Organization
,	O
and	O
the	O
7th	O
Fateh	O
Conference	O
of	O
the	O
Palestinian	B-Organization
National	I-Organization
Liberation	I-Organization
Front	I-Organization
(	O
held	O
in	O
late	O
2016	O
)	O
.	O

The	O
titles	O
and	O
contents	O
of	O
these	O
files	O
suggest	O
that	O
the	O
actor	O
targeted	O
individuals	O
affiliated	O
with	O
these	O
government	O
agencies	O
and	O
the	O
Fatah	B-Organization
political	O
party	O
.	O

Some	O
malicious	O
files	O
associated	O
with	O
these	O
samples	O
were	O
titled	O
the	O
following	O
:	O
Council_of_ministres_decision	O
Minutes	O
of	O
the	O
Geneva	O
Meeting	O
on	O
Troops	O
Summary	O
of	O
today	O
's	O
meetings.doc.exe	B-Indicator
The	O
most	O
important	O
points	O
of	O
meeting	O
the	O
memory	O
of	O
the	O
late	O
President	O
Abu	O
Omar	O
may	O
Allah	O
have	O
mercy	O
on	O
him	O
-	O
Paper	O
No	O
.	O

1	O
Fadi	O
Alsalamin	O
scandal	O
with	O
an	O
Israeli	O
officer	O
-	O
exclusive	O
-	O
watched	O
before	O
the	O
deletion	O
-	O
Fadi	O
Elsalameen	O
The	O
details	O
of	O
the	O
assassination	O
of	O
President	O
Arafat_06-12-2016_docx	O
Quds.rar	B-Indicator
Many	O
of	O
these	O
executables	O
are	O
associated	O
with	O
various	O
short	O
links	O
created	O
using	O
Bit.ly	B-System
,	O
a	O
URL	O
shortening	O
service	O
.	O

After	O
analyzing	O
the	O
traffic	O
associated	O
with	O
these	O
short	O
links	O
,	O
we	O
determined	O
that	O
each	O
one	O
was	O
associated	O
with	O
a	O
referral	O
path	O
from	O
mail.mosa.pna.ps	B-Indicator
.	O

MOSA	B-Organization
is	O
the	O
Palestinian	O
Directorate	O
of	O
Social	O
Development	O
whose	O
mandate	O
is	O
to	O
achieve	O
comprehensive	O
development	O
,	O
social	O
security	O
,	O
and	O
economic	O
growth	O
for	O
Palestinian	O
families	O
,	O
according	O
to	O
publicly	O
available	O
information	O
on	O
this	O
ministry	O
.	O

Infrastructure	O
At	O
the	O
time	O
of	O
writing	O
the	O
following	O
domains	O
have	O
either	O
been	O
used	O
by	O
this	O
family	O
or	O
are	O
currently	O
active	O
.	O

We	O
expect	O
this	O
list	O
to	O
grow	O
given	O
that	O
this	O
actor	O
has	O
changed	O
its	O
infrastructure	O
numerous	O
times	O
in	O
2017	O
.	O

cecilia-gilbert	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
comgooogel	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
orgmary-crawley	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
commydriveweb	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
comrose-sturat	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
infokalisi	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
xyzdebra-morgan	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
comarnani	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
infoacount-manager	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
infogooogel-drive	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
commediauploader	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
meacount-manager	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
netupload404	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
clubupload999	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
infoal-amalhumandevelopment	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
commargaery	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
coupload202	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
comgo-mail-accounts	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
comupload101	I-Indicator
[	B-Indicator
.	I-Indicator

]	I-Indicator
netsybil-parks	I-Indicator
[	O
.	O

]	O
infodavos-seaworth	O
[	O
.	O

]	O
infoupload999	O
[	O
.	O

]	O
orgacount-manager	O
[	O
.	O

]	O
comlila-tournai	O
[	O
.	O

]	O
comaccount-manager	O
[	O
.	O

]	O
orgmediauploader	O
[	O
.	O

]	O
infokalisi	O
[	O
.	O

]	O
orgaryastark	O
[	O
.	O

]	O
infomavis-dracula	O
[	O
.	O

]	O
comkalisi	O
[	O
.	O

]	O
infogoogle-support-team	O
[	O
.	O

]	O
com9oo91e	O
[	O
.	O

]	O
comuseraccount	O
[	O
.	O

]	O
websiteaccounts-fb	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
comakashipro	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
comfeteh-asefa	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
comlagertha-lothbrok	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
info	I-Indicator
OpSec	O
fails	O
and	O
use	O
of	O
cryptography	O
While	O
looking	O
at	O
this	O
infrastructure	O
,	O
we	O
identified	O
that	O
one	O
of	O
these	O
domains	O
has	O
directory	O
indexing	O
enabled	O
.	O

This	O
mistake	O
in	O
operational	O
security	O
allowed	O
us	O
to	O
gain	O
visibility	O
into	O
exfiltrated	O
content	O
for	O
a	O
number	O
of	O
devices	O
.	O

Continued	O
mirroring	O
suggests	O
it	O
is	O
likely	O
a	O
regularly	O
cleaned	O
staging	O
server	O
.	O

We	O
sourced	O
the	O
over	O
561MB	O
of	O
exfiltrated	O
data	O
from	O
this	O
domain	O
alone	O
,	O
all	O
of	O
which	O
we	O
found	O
to	O
be	O
7z	O
compressed	O
and	O
password	O
protected	O
.	O

Password	O
generation	O
for	O
compressed	O
files	O
takes	O
place	O
client-side	O
with	O
each	O
device	O
using	O
a	O
unique	O
key	O
in	O
most	O
scenarios	O
.	O

Key	O
information	O
consists	O
of	O
an	O
MD5	O
hash	O
of	O
the	O
device	O
's	O
Android	B-System
ID	O
,	O
the	O
device	O
manufacturer	O
,	O
and	O
the	O
device	O
model	O
with	O
each	O
separated	O
by	O
an	O
underscore	O
.	O

Visually	O
,	O
this	O
can	O
be	O
represented	O
as	O
follows	O
:	O
Android	B-System
ID	O
When	O
combined	O
with	O
our	O
analysis	O
of	O
indexed	O
directories	O
on	O
C2	O
infrastructure	O
,	O
we	O
were	O
able	O
to	O
easily	O
automate	O
the	O
generation	O
of	O
the	O
password	O
used	O
by	O
each	O
device	O
and	O
,	O
in	O
turn	O
,	O
successfully	O
decompress	O
all	O
exfiltrated	O
content	O
from	O
compromised	O
devices	O
.	O

Indexed	O
directories	O
on	O
C2	O
infrastructure	O
While	O
exfiltrated	O
content	O
is	O
encrypted	O
,	O
information	O
used	O
to	O
generate	O
the	O
password	O
is	O
plainly	O
visible	O
in	O
the	O
top	O
level	O
directories	O
for	O
each	O
device	O
.	O

Taking	O
this	O
information	O
from	O
directory	O
listings	O
,	O
like	O
the	O
one	O
shown	O
above	O
,	O
allowed	O
for	O
the	O
decryption	O
of	O
all	O
content	O
.	O

In	O
this	O
case	O
,	O
FrozenCell	B-Malware
has	O
primarily	O
netted	O
the	O
actors	O
behind	O
it	O
with	O
recorded	O
outbound	O
calls	O
followed	O
closely	O
by	O
images	O
and	O
recorded	O
incoming	O
calls	O
.	O

FrozenCell	B-Malware
is	O
part	O
of	O
a	O
very	O
successful	O
,	O
multi-platform	O
surveillance	O
campaign	O
.	O

Attackers	O
are	O
growing	O
smarter	O
,	O
targeting	O
individuals	O
through	O
the	O
devices	O
and	O
the	O
services	O
they	O
use	O
most	O
.	O

Government	O
agencies	O
and	O
enterprises	O
should	O
plan	O
to	O
be	O
hit	O
from	O
all	O
angles	O
-	O
cloud	O
services	O
,	O
mobile	O
devices	O
,	O
laptops	O
-	O
in	O
order	O
to	O
build	O
comprehensive	O
security	O
strategies	O
that	O
work	O
.	O

TUESDAY	O
,	O
MAY	O
19	O
,	O
2020	O
The	O
wolf	O
is	O
back	O
...	O
NEWS	O
SUMMARY	O
Thai	O
Android	B-System
devices	O
and	O
users	O
are	O
being	O
targeted	O
by	O
a	O
modified	O
version	O
of	O
DenDroid	B-Malware
we	O
are	O
calling	O
"	O
WolfRAT	B-Malware
,	O
''	O
now	O
targeting	O
messaging	O
apps	O
like	O
WhatsApp	B-System
,	O
Facebook	B-System
Messenger	I-System
and	O
Line	B-System
.	O

We	O
assess	O
with	O
high	O
confidence	O
that	O
this	O
modified	O
version	O
is	O
operated	O
by	O
the	O
infamous	O
Wolf	B-Organization
Research	I-Organization
.	O

This	O
actor	O
has	O
shown	O
a	O
surprising	O
level	O
of	O
amateur	O
actions	O
,	O
including	O
code	O
overlaps	O
,	O
open-source	O
project	O
copy/paste	O
,	O
classes	O
never	O
being	O
instanced	O
,	O
unstable	O
packages	O
and	O
unsecured	O
panels	O
.	O

EXECUTIVE	O
SUMMARY	O
Cisco	B-Organization
Talos	I-Organization
has	O
discovered	O
a	O
new	O
Android	O
malware	O
based	O
on	O
a	O
leak	O
of	O
the	O
DenDroid	B-Malware
malware	O
family	O
.	O

We	O
named	O
this	O
malware	O
"	O
WolfRAT	B-Malware
''	O
due	O
to	O
strong	O
links	O
between	O
this	O
malware	O
(	O
and	O
the	O
command	O
and	O
control	O
(	O
C2	O
)	O
infrastructure	O
)	O
and	O
Wolf	B-Organization
Research	I-Organization
,	O
an	O
infamous	O
organization	O
that	O
developed	O
interception	O
and	O
espionage-based	O
malware	O
and	O
was	O
publicly	O
described	O
by	O
CSIS	O
during	O
Virus	O
Bulletin	O
2018	O
.	O

We	O
identified	O
infrastructure	O
overlaps	O
and	O
string	O
references	O
to	O
previous	O
Wolf	B-Organization
Research	I-Organization
work	O
.	O

The	O
organization	O
appears	O
to	O
be	O
shut	O
down	O
,	O
but	O
the	O
threat	O
actors	O
are	O
still	O
very	O
active	O
.	O

We	O
identified	O
campaigns	O
targeting	O
Thai	O
users	O
and	O
their	O
devices	O
.	O

Some	O
of	O
the	O
C2	O
servers	O
are	O
located	O
in	O
Thailand	O
.	O

The	O
panels	O
also	O
contain	O
Thai	O
JavaScript	O
comments	O
and	O
the	O
domain	O
names	O
also	O
contain	O
references	O
to	O
Thai	O
food	O
,	O
a	O
tactic	O
commonly	O
employed	O
to	O
entice	O
users	O
to	O
click/visit	O
these	O
C2	O
panels	O
without	O
much	O
disruption	O
.	O

We	O
identified	O
a	O
notable	O
lack	O
of	O
sophistication	O
in	O
this	O
investigation	O
such	O
as	O
copy/paste	O
,	O
unstable	O
code	O
,	O
dead	O
code	O
and	O
panels	O
that	O
are	O
freely	O
open	O
.	O

What	O
's	O
new	O
?	O

WolfRAT	B-Malware
is	O
based	O
on	O
a	O
previously	O
leaked	O
malware	O
named	O
DenDroid	B-Malware
.	O

The	O
new	O
malware	O
appears	O
to	O
be	O
linked	O
to	O
the	O
infamous	O
Wolf	B-Organization
Research	I-Organization
organization	O
and	O
targets	O
Android	B-System
devices	O
located	O
in	O
Thailand	O
.	O

How	O
did	O
it	O
work	O
?	O

The	O
malware	O
mimics	O
legit	O
services	O
such	O
as	O
Google	B-Organization
service	O
,	O
GooglePlay	B-System
or	O
Flash	B-System
update	O
.	O

The	O
malware	O
is	O
not	O
really	O
advanced	O
and	O
is	O
based	O
on	O
a	O
lot	O
of	O
copy/paste	O
from	O
public	O
sources	O
available	O
on	O
the	O
Internet	O
.	O

The	O
C2	O
infrastructure	O
contains	O
a	O
lack	O
of	O
sophistication	O
such	O
as	O
open	O
panels	O
,	O
reuse	O
of	O
old	O
servers	O
publicly	O
tagged	O
as	O
malicious…	O
So	O
what	O
?	O

After	O
being	O
publicly	O
denounced	O
by	O
CSIS	B-Organization
Group	I-Organization
—	O
a	O
threat	O
intelligence	O
company	O
in	O
Denmark	O
—	O
Wolf	B-Organization
Research	I-Organization
was	O
closed	O
and	O
a	O
new	O
organization	O
named	O
LokD	B-Organization
was	O
created	O
.	O

This	O
new	O
organization	O
seems	O
to	O
work	O
on	O
securing	O
Android	B-Organization
devices	O
.	O

However	O
,	O
thanks	O
to	O
the	O
infrastructure	O
sharing	O
and	O
forgotten	O
panel	O
names	O
,	O
we	O
assess	O
with	O
high	O
confidence	O
that	O
this	O
actor	O
is	O
still	O
active	O
,	O
it	O
is	O
still	O
developing	O
malware	O
and	O
has	O
been	O
using	O
it	O
from	O
mid-June	O
to	O
today	O
.	O

On	O
the	O
C2	O
panel	O
,	O
we	O
found	O
a	O
potential	O
link	O
between	O
Wolf	B-Organization
Research	I-Organization
and	O
another	O
Cyprus	O
organization	O
named	O
Coralco	B-Organization
Tech	I-Organization
.	O

This	O
organization	O
is	O
also	O
working	O
on	O
interception	O
technology	O
.	O

LINKS	O
TO	O
WOLF	O
INTELLIGENCE	O
During	O
the	O
Virus	O
Bulletin	O
conference	O
in	O
2018	O
,	O
CSIS	B-Organization
researchers	O
Benoît	O
Ancel	O
and	O
Aleksejs	O
Kuprins	O
did	O
a	O
presentation	O
on	O
Wolf	B-Organization
Research	I-Organization
and	O
the	O
offensive	O
arsenal	O
developed	O
by	O
the	O
organization	O
.	O

They	O
mentioned	O
an	O
Android	B-System
,	O
iOS	B-System
and	O
Windows	B-System
remote	O
access	O
tool	O
(	O
RAT	O
)	O
.	O

Their	O
findings	O
showed	O
that	O
Wolf	O
is	O
headquartered	O
in	O
Germany	O
with	O
offices	O
in	O
Cyprus	O
,	O
Bulgaria	O
,	O
Romania	O
,	O
India	O
and	O
(	O
possibly	O
)	O
the	O
U.S	O
.	O

The	O
organization	O
was	O
closed	O
after	O
the	O
CSIS	B-Organization
presentation	O
.	O

However	O
,	O
the	O
director	O
created	O
a	O
new	O
organization	O
in	O
Cyprus	O
named	O
LokD	B-Organization
.	O

This	O
new	O
organization	O
proposed	O
the	O
creation	O
of	O
a	O
more	O
secure	O
Android	B-System
phone	O
.	O

Based	O
on	O
the	O
organization	O
website	O
,	O
it	O
also	O
proposes	O
services	O
and	O
developed	O
zero-day	B-Vulnerability
vulnerabilities	I-Vulnerability
to	O
test	O
their	O
own	O
products	O
:	O
Zero-day	O
research	O
from	O
lokd.com	B-Organization
We	O
can	O
see	O
that	O
the	O
organization	O
owner	O
still	O
has	O
an	O
interest	O
in	O
Android	B-System
devices	O
.	O

Based	O
on	O
infrastructure	O
overlaps	O
and	O
leaked	O
information	O
,	O
we	O
assess	O
with	O
high	O
confidence	O
that	O
the	O
malware	O
we	O
identified	O
and	O
present	O
in	O
this	O
paper	O
is	O
linked	O
to	O
Wolf	B-Organization
Research	I-Organization
.	O

One	O
of	O
the	O
samples	O
(	O
e19823a1ba4a0e40cf459f4a0489fc257720cc0d71ecfb7ad94b3ca86fbd85d1	B-Indicator
)	O
uses	O
the	O
C2	O
server	O
svcws	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ponethus	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
.	O

Based	O
on	O
our	O
research	O
and	O
Benoît	O
Ancel	O
's	O
tracker	O
,	O
this	O
C2	O
was	O
used	O
by	O
Wolf	B-Organization
Intelligence	I-Organization
:	O
Additionally	O
,	O
we	O
identified	O
two	O
empty	O
panels	O
on	O
a	O
C2	O
server	O
.	O

The	O
new	O
one	O
with	O
the	O
title	O
"	O
Coralco	O
Archimedes	O
,	O
''	O
and	O
an	O
older	O
version	O
with	O
the	O
title	O
"	O
Wolf	O
Intelligence	O
:	O
''	O
New	O
panel	O
Old	O
panel	O
The	O
new	O
panel	O
name	O
contains	O
"	O
Coralco	O
''	O
in	O
its	O
name	O
.	O

Coralco	B-Organization
Tech	I-Organization
is	O
an	O
organization	O
located	O
in	O
Cyprus	O
and	O
providing	O
interception	O
tools	O
.	O

We	O
can	O
not	O
say	O
for	O
sure	O
if	O
Wolf	B-Organization
Research	I-Organization
and	O
Coralco	B-Organization
Tech	I-Organization
are	O
linked	O
,	O
but	O
this	O
panel	O
name	O
,	O
their	O
offerings	O
and	O
the	O
panel	O
layout	O
would	O
suggest	O
it	O
should	O
be	O
considered	O
suspiciously	O
linked	O
.	O

Coralco	O
Tech	O
's	O
services	O
description	O
.	O

VICTIMOLOGY	O
ON	O
THE	O
IDENTIFIED	O
CAMPAIGNS	O
The	O
campaigns	O
we	O
analyzed	O
targeted	O
Android	B-System
devices	O
in	O
Thailand	O
.	O

The	O
C2	O
server	O
domain	O
is	O
linked	O
to	O
Thai	O
food	O
:	O
Nampriknum	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
net	I-Indicator
:	O
Nam	O
Phrik	O
Num	O
Somtum	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
:	O
Som	O
Tum	O
We	O
also	O
identified	O
comments	O
in	O
Thai	O
on	O
the	O
C2	O
infrastructure	O
mentioned	O
in	O
the	O
previous	O
chapter	O
:	O
MALWARE	O
DenDroid	B-Malware
The	O
Android	B-System
malware	O
is	O
based	O
on	O
the	O
DenDroid	B-Malware
Android	O
malware	O
.	O

Several	O
analysis	O
reports	O
were	O
published	O
on	O
this	O
malware	O
in	O
2014	O
and	O
,	O
finally	O
,	O
the	O
source	O
code	O
was	O
leaked	O
in	O
2015	O
.	O

The	O
original	O
leak	O
is	O
no	O
longer	O
available	O
on	O
github.com	O
,	O
but	O
a	O
copy	O
can	O
be	O
found	O
here	O
.	O

The	O
table	O
below	O
shows	O
the	O
commands	O
available	O
to	O
the	O
operator	O
for	O
tasking	O
on	O
infected	O
devices	O
.	O

This	O
malware	O
is	O
simplistic	O
in	O
comparison	O
to	O
some	O
modern-day	O
Android	B-System
malware	O
.	O

The	O
best	O
example	O
of	O
that	O
is	O
that	O
it	O
does	O
n't	O
take	O
advantage	O
of	O
the	O
accessibility	O
framework	O
,	O
collecting	O
information	O
on	O
non-rooted	O
devices	O
.	O

The	O
commands	O
are	O
self-explanatory	O
and	O
show	O
the	O
features	O
included	O
in	O
the	O
malware	O
.	O

Some	O
of	O
them	O
like	O
takephoto	O
,	O
takevideo	O
,	O
recordaudio	O
,	O
getsentsms	O
and	O
uploadpictures	O
are	O
focused	O
on	O
espionage	O
activities	O
.	O

Others	O
like	O
transferbot	O
,	O
promptupdate	O
and	O
promptuninstall	O
are	O
meant	O
to	O
help	O
the	O
operator	O
manage	O
the	O
malware	O
.	O

Version	O
#	O
1	O
:	O
June	O
2019	O
—	O
Domain	O
:	O
databit	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
During	O
our	O
investigation	O
,	O
we	O
identified	O
at	O
least	O
four	O
major	O
releases	O
of	O
the	O
RAT	O
.	O

The	O
permissions	O
on	O
the	O
first	O
version	O
of	O
the	O
malware	O
lay	O
out	O
the	O
foundations	O
of	O
a	O
spying	O
trojan	O
.	O

Permissions	O
The	O
package	O
name	O
follows	O
the	O
original	O
style	O
name	O
used	O
on	O
DenDroid	B-Malware
.	O

The	O
code	O
is	O
obfuscated	O
but	O
not	O
packed	O
.	O

This	O
malware	O
also	O
contains	O
a	O
screen	O
recorder	O
.	O

This	O
feature	O
is	O
implemented	O
using	O
another	O
open-source	O
software	O
package	O
that	O
can	O
be	O
found	O
here	O
.	O

The	O
service	O
is	O
implemented	O
in	O
the	O
class	O
com.serenegiant.service.ScreenRecorderService	B-Indicator
which	O
is	O
declared	O
in	O
the	O
package	O
manifest	O
.	O

During	O
our	O
analysis	O
of	O
this	O
sample	O
,	O
we	O
did	O
notice	O
that	O
the	O
class	O
itself	O
is	O
never	O
called	O
or	O
used	O
by	O
the	O
malware	O
.	O

It	O
remains	O
available	O
within	O
the	O
source	O
code	O
but	O
no	O
method	O
of	O
use	O
takes	O
place	O
.	O

Version	O
#	O
2	O
:	O
June	O
-	O
Aug.	O
2019	O
—	O
Domain	O
:	O
somtum	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
This	O
is	O
the	O
first	O
version	O
that	O
shows	O
the	O
code	O
organization	O
evolution	O
that	O
will	O
continue	O
to	O
be	O
used	O
on	O
all	O
other	O
functions	O
throughout	O
this	O
malware	O
.	O

Code	O
structure	O
Obviously	O
,	O
this	O
code	O
is	O
not	O
obfuscated	O
when	O
compared	O
with	O
the	O
previous	O
version	O
it	O
becomes	O
clear	O
that	O
this	O
is	O
the	O
same	O
code	O
base	O
.	O

One	O
of	O
the	O
first	O
changes	O
that	O
stands	O
out	O
is	O
that	O
the	O
screen	O
recording	O
feature	O
mentioned	O
in	O
the	O
previous	O
sample	O
has	O
been	O
removed	O
.	O

A	O
new	O
class	O
was	O
added	O
called	O
com.utils.RestClient	B-Indicator
.	O

This	O
class	O
is	O
based	O
on	O
public	O
code	O
belonging	O
to	O
the	O
package	O
praeda.muzikmekan	B-Indicator
,	O
which	O
can	O
be	O
found	O
here	O
among	O
other	O
places	O
.	O

Just	O
like	O
in	O
previous	O
examples	O
,	O
the	O
malware	O
author	O
does	O
not	O
use	O
this	O
package	O
.	O

Missing	O
permissions	O
The	O
lack	O
of	O
the	O
READ_FRAME_BUFFER	O
permission	O
can	O
be	O
justified	O
by	O
the	O
removal	O
of	O
the	O
screen	O
record	O
feature	O
.	O

The	O
ACCESS_SUPERUSER	O
may	O
have	O
been	O
removed	O
because	O
it	O
was	O
deprecated	O
upon	O
the	O
release	O
of	O
Android	B-System
5.0	I-System
Lollipop	B-System
which	O
happened	O
in	O
2014	O
.	O

The	O
reality	O
is	O
that	O
the	O
RAT	O
permissions	O
can	O
be	O
implemented	O
just	O
with	O
the	O
permissions	O
declared	O
on	O
the	O
manifest	O
,	O
thus	O
there	O
is	O
no	O
need	O
for	O
higher	O
permissions	O
.	O

Version	O
#	O
3	O
:	O
Sept.	O
-	O
Dec.	O
2019	O
—	O
Domain	O
:	O
ponethus	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
Given	O
that	O
there	O
is	O
some	O
overlap	O
in	O
the	O
previous	O
two	O
versions	O
,	O
it	O
came	O
as	O
no	O
surprise	O
to	O
us	O
that	O
we	O
finally	O
identified	O
a	O
sample	O
which	O
is	O
an	O
evolution	O
based	O
on	O
both	O
previous	O
versions	O
.	O

This	O
sample	O
is	O
clearly	O
a	O
mix	O
between	O
the	O
two	O
.	O

This	O
is	O
also	O
the	O
first	O
version	O
where	O
the	O
package	O
name	O
changes	O
into	O
something	O
that	O
a	O
less	O
aware	O
user	O
may	O
be	O
tricked	O
by	O
,	O
com.android.playup	B-Indicator
.	O

This	O
version	O
brings	O
back	O
the	O
ACCESS_SUPERUSER	O
and	O
READ_FRAME_BUFFER	O
permissions	O
.	O

However	O
,	O
this	O
time	O
,	O
the	O
permission	O
is	O
actually	O
used	O
.	O

WhatsApp	B-System
message	O
capture	O
The	O
service	O
com.serenegiant.service.ScreenRecorderService	B-Indicator
,	O
is	O
invoked	O
by	O
the	O
ScreenRecorderActivity	O
.	O

Upon	O
creation	O
,	O
this	O
activity	O
launches	O
a	O
thread	O
that	O
will	O
loop	O
on	O
a	O
50-second	O
interval	O
.	O

In	O
the	O
first	O
iteration	O
,	O
the	O
screen	O
recording	O
is	O
started	O
and	O
will	O
only	O
stop	O
when	O
the	O
RAT	O
determines	O
that	O
WhatsApp	B-System
is	O
not	O
running	O
.	O

It	O
's	O
restarted	O
in	O
the	O
next	O
cycle	O
independently	O
based	O
on	O
if	O
WhatsApp	B-System
is	O
running	O
.	O

In	O
this	O
version	O
,	O
the	O
developer	O
added	O
more	O
classes	O
from	O
the	O
same	O
package	O
.	O

Even	O
though	O
we	O
could	O
not	O
find	O
indications	O
of	O
being	O
in	O
use	O
,	O
two	O
stand	O
out	O
.	O

Bluetooth	O
—	O
which	O
allows	O
the	O
interaction	O
with	O
the	O
Bluetooth	O
interface	O
,	O
and	O
net/deacon	O
—	O
which	O
implements	O
a	O
beaconing	O
system	O
based	O
on	O
UDP	O
.	O

Android	B-System
shell	O
A	O
new	O
package	O
was	O
added	O
that	O
allows	O
the	O
execution	O
of	O
commands	O
in	O
the	O
Android	B-System
shell	O
.	O

Again	O
,	O
this	O
package	O
source	O
code	O
is	O
publicly	O
available	O
and	O
can	O
be	O
found	O
here	O
.	O

One	O
of	O
the	O
uses	O
the	O
malware	O
gives	O
to	O
this	O
package	O
is	O
the	O
execution	O
of	O
the	O
command	O
"	O
dumpsys	O
''	O
to	O
determine	O
if	O
certain	O
activities	O
are	O
running	O
.	O

Check	O
if	O
chat	O
apps	O
are	O
running	O
In	O
the	O
above	O
example	O
,	O
the	O
malware	O
is	O
searching	O
for	O
Line	O
,	O
Facebook	B-System
Messenger	I-System
and	O
WhatsApp	B-System
activities	O
.	O

This	O
is	O
part	O
of	O
a	O
class	O
called	O
CaptureService	O
,	O
which	O
already	O
existed	O
in	O
the	O
previous	O
version	O
but	O
it	O
was	O
not	O
duly	O
implemented	O
.	O

Previous	O
version	O
The	O
capture	O
service	O
class	O
implements	O
the	O
chat	O
applications	O
interception	O
.	O

Upon	O
creation	O
the	O
class	O
will	O
start	O
to	O
take	O
screenshots	O
that	O
will	O
be	O
stopped	O
and	O
uploaded	O
to	O
the	O
C2	O
once	O
the	O
service	O
ca	O
n't	O
find	O
the	O
targeted	O
applications	O
running	O
.	O

The	O
core	O
of	O
this	O
functionality	O
is	O
also	O
based	O
on	O
an	O
open-source	O
project	O
that	O
can	O
be	O
found	O
here	O
.	O

Another	O
novelty	O
is	O
a	O
VPN-related	O
package	O
,	O
which	O
is	O
based	O
on	O
OrbotVPN	B-System
.	O

Once	O
again	O
,	O
it	O
does	O
n't	O
seem	O
to	O
actually	O
be	O
in	O
use	O
.	O

The	O
same	O
happens	O
with	O
the	O
package	O
squareup.otto	B-Indicator
,	O
which	O
is	O
an	O
open-source	O
bus	O
implementation	O
focused	O
on	O
Android	B-System
implementation	O
.	O

Both	O
sources	O
can	O
be	O
found	O
here	O
and	O
here	O
.	O

Version	O
#	O
4	O
:	O
April	O
2020	O
—	O
Domain	O
:	O
nampriknum.net	B-Indicator
Following	O
the	O
same	O
pattern	O
,	O
this	O
version	O
has	O
some	O
added	O
features	O
and	O
others	O
,	O
which	O
were	O
not	O
in	O
use	O
,	O
removed	O
.	O

First	O
of	O
all	O
the	O
new	O
package	O
name	O
is	O
com.google.services	B-Indicator
,	O
which	O
can	O
easily	O
be	O
confused	O
with	O
a	O
legitimate	O
Google	B-Organization
service	O
.	O

The	O
VPN	O
package	O
is	O
no	O
longer	O
present	O
,	O
further	O
reinforcing	O
our	O
conclusion	O
that	O
it	O
was	O
not	O
in	O
use	O
.	O

WolfRAT	B-Malware
application	O
screen	O
The	O
Google	B-System
GMS	I-System
and	O
Firebase	B-System
service	O
has	O
been	O
added	O
,	O
however	O
,	O
no	O
configuration	O
has	O
been	O
found	O
,	O
even	O
though	O
services	O
seem	O
to	O
be	O
referenced	O
in	O
the	O
of	O
a	O
new	O
class	O
.	O

The	O
new	O
class	O
is	O
called	O
NotificationListener	O
and	O
extends	O
the	O
NotificationListenerService	O
class	O
.	O

This	O
would	O
allow	O
the	O
RAT	O
to	O
receive	O
system	O
notifications	O
.	O

Notification	O
handling	O
method	O
The	O
class	O
is	O
only	O
implemented	O
in	O
debug	O
mode	O
,	O
pushing	O
all	O
captured	O
information	O
into	O
the	O
log	O
.	O

The	O
usage	O
of	O
the	O
PlusShare	B-System
API	O
in	O
2020	O
denotes	O
some	O
unprofessional	O
development	O
,	O
since	O
this	O
is	O
the	O
API	O
to	O
access	O
Google+	B-Organization
.	O

This	O
service	O
,	O
along	O
with	O
the	O
API	O
,	O
was	O
fully	O
decommissioned	O
in	O
March	O
2019	O
.	O

This	O
version	O
adds	O
one	O
significant	O
class	O
—	O
it	O
requests	O
DEVICE_ADMIN	O
privileges	O
.	O

Device	O
admin	O
policies	O
Looking	O
at	O
the	O
policy	O
's	O
definition	O
,	O
we	O
can	O
see	O
that	O
it	O
lists	O
all	O
the	O
available	O
policies	O
even	O
if	O
most	O
of	O
them	O
are	O
deprecated	O
on	O
Android	B-System
10.0	I-System
and	O
their	O
usage	O
results	O
in	O
a	O
security	O
exception	O
.	O

The	O
code	O
implementation	O
again	O
seems	O
that	O
it	O
has	O
been	O
added	O
for	O
testing	O
purposes	O
only	O
.	O

Versions	O
overview	O
The	O
DenDroid	B-Malware
code	O
base	O
was	O
kept	O
to	O
such	O
an	O
extent	O
that	O
even	O
the	O
original	O
base64-encoded	O
password	O
was	O
kept	O
.	O

Original	O
password	O
The	O
main	O
service	O
follows	O
the	O
same	O
structure	O
as	O
the	O
first	O
version	O
,	O
the	O
anti-analysis	O
features	O
are	O
primitive	O
,	O
only	O
checking	O
the	O
emulator	O
environment	O
without	O
any	O
kind	O
of	O
packing	O
or	O
obfuscation	O
.	O

The	O
malware	O
will	O
start	O
the	O
main	O
service	O
if	O
all	O
the	O
requested	O
permissions	O
and	O
the	O
device	O
admin	O
privileges	O
are	O
granted	O
.	O

Otherwise	O
,	O
it	O
will	O
launch	O
an	O
ACTION_APPLICATION_SETTINGS	O
intent	O
trying	O
to	O
trick	O
the	O
user	O
to	O
grant	O
the	O
permissions	O
.	O

Each	O
sample	O
contains	O
a	O
userId	O
hardcoded	O
,	O
meaning	O
that	O
each	O
sample	O
can	O
only	O
be	O
used	O
in	O
a	O
victim	O
.	O

It	O
seems	O
,	O
however	O
,	O
if	O
the	O
same	O
victim	O
has	O
more	O
than	O
one	O
device	O
the	O
malware	O
can	O
be	O
reused	O
since	O
the	O
IMEI	O
is	O
sent	O
along	O
with	O
each	O
data	O
exfiltration	O
.	O

It	O
is	O
clear	O
that	O
this	O
RAT	O
is	O
under	O
intense	O
development	O
,	O
however	O
,	O
the	O
addition	O
and	O
removal	O
of	O
packages	O
,	O
along	O
with	O
the	O
huge	O
quantity	O
of	O
unused	O
code	O
and	O
usage	O
of	O
deprecated	O
and	O
old	O
techniques	O
denotes	O
an	O
amateur	O
development	O
methodology	O
.	O

CONCLUSION	O
We	O
witness	O
actors	O
continually	O
using	O
open-source	O
platforms	O
,	O
code	O
and	O
packages	O
to	O
create	O
their	O
own	O
software	O
.	O

Some	O
are	O
carried	O
out	O
well	O
,	O
others	O
,	O
like	O
WolfRAT	B-Malware
,	O
are	O
designed	O
with	O
an	O
overload	O
of	O
functionality	O
in	O
mind	O
as	O
opposed	O
to	O
factoring	O
any	O
sensible	O
approach	O
to	O
the	O
development	O
aspect	O
.	O

After	O
all	O
,	O
a	O
working	O
product	O
is	O
often	O
more	O
important	O
than	O
a	O
stable	O
product	O
.	O

We	O
watched	O
WolfRAT	B-Malware
evolve	O
through	O
various	O
iterations	O
which	O
shows	O
that	O
the	O
actor	O
wanted	O
to	O
ensure	O
functional	O
improvements	O
—	O
perhaps	O
they	O
had	O
deadlines	O
to	O
meet	O
for	O
their	O
customers	O
,	O
but	O
with	O
no	O
thought	O
given	O
to	O
removing	O
old	O
code	O
blocks	O
,	O
classes	O
,	O
etc	O
.	O

throughout	O
the	O
Android	B-System
package	O
.	O

WolfRAT	B-Malware
is	O
a	O
specifically	O
targeted	O
RAT	O
which	O
we	O
assess	O
to	O
be	O
aimed	O
at	O
Thai	O
individuals	O
and	O
,	O
based	O
on	O
previous	O
work	O
from	O
Wolf	B-Organization
Research	I-Organization
,	I-Organization
most	O
likely	O
used	O
as	O
an	O
intelligence-gathering	O
tool	O
or	O
interception	O
tool	O
.	O

This	O
can	O
be	O
packaged	O
and	O
"	O
sold	O
''	O
in	O
many	O
different	O
ways	O
to	O
customers	O
.	O

A	O
"	O
Tracking	O
tool	O
''	O
or	O
an	O
"	O
Admin	O
tool	O
''	O
are	O
often	O
cited	O
for	O
these	O
kinds	O
of	O
tools	O
for	O
"	O
commercial	O
''	O
or	O
"	O
enterprise	O
''	O
usage	O
.	O

Wolf	I-Malware
Research	I-Organization
claimed	O
to	O
shut	O
down	O
their	O
operations	O
but	O
we	O
clearly	O
see	O
that	O
their	O
previous	O
work	O
continues	O
under	O
another	O
guise	O
.	O

The	O
ability	O
to	O
carry	O
out	O
these	O
types	O
of	O
intelligence-gathering	O
activities	O
on	O
phones	O
represents	O
a	O
huge	O
score	O
for	O
the	O
operator	O
.	O

The	O
chat	O
details	O
,	O
WhatsApp	B-System
records	O
,	O
messengers	O
and	O
SMSs	O
of	O
the	O
world	O
carry	O
some	O
sensitive	O
information	O
which	O
people	O
often	O
forget	O
when	O
communicating	O
with	O
their	O
devices	O
.	O

We	O
see	O
WolfRAT	B-Malware
specifically	O
targeting	O
a	O
highly	O
popular	O
encrypted	O
chat	O
app	O
in	O
Asia	O
,	O
Line	B-System
,	O
which	O
suggests	O
that	O
even	O
a	O
careful	O
user	O
with	O
some	O
awareness	O
around	O
end-to-end	O
encryption	O
chats	O
would	O
still	O
be	O
at	O
the	O
mercy	O
of	O
WolfRAT	B-Malware
and	O
it	O
's	O
prying	O
eyes	O
.	O

IOCS	O
Hashes	O
139edb1bc033725539b117f50786f3d3362ed45845c57fe1f82e7ed72b044367	B-Indicator
e19823a1ba4a0e40cf459f4a0489fc257720cc0d71ecfb7ad94b3ca86fbd85d1	B-Indicator
e19823a1ba4a0e40cf459f4a0489fc257720cc0d71ecfb7ad94b3ca86fbd85d1	B-Indicator
e5f346d8f312cc1f93c2c6af611e2f50805c528934786ea173cabc6a39b14cda	B-Indicator

1849a50a6ac9b3eec51492745eeb14765fe2e78488d476b0336d8e41c2c581d4	B-Indicator
d328fca14c4340fcd4a15e47562a436085e6b1bb5376b5ebd83d3e7218db64e7	B-Indicator
59b9809dba857c5969f23f460a2bf0a337a71622a79671066675ec0acf89c810	B-Indicator
120474682ea439eb0b28274c495d9610a73d892a4b8feeff268c670570db97e2	B-Indicator

ed234e61849dcb95223676abe2312e1378d6130c0b00851d82cda545b946ec83	B-Indicator
27410d4019251a70d38f0635277f931fb73f67ac9f2e1f3b475ce680ebfde12a	B-Indicator
6e6c210535b414c5aa2dd9e67f5153feeb43a8ac8126d8e249e768f501323a3e	B-Indicator
4a32ced20df7001da7d29edc31ca76e13eef0c9b355f62c44888853435e9794f	B-Indicator

ac5abaebd9f516b8b389450f7d27649801d746fb14963b848f9d6dad0a505e66	B-Indicator
3a45d7a16937d4108b5b48f44d72bb319be645cbe15f003dc9e77fd52f45c065	B-Indicator
Domains	O
cvcws	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ponethus	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
svc	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ponethus	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
www	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ponethus	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
webmail	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ponethus	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
com	I-Indicator
nampriknum	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
net	I-Indicator
www	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
nampriknum	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
net	I-Indicator
svc	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
nampriknum	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
net	I-Indicator
svcws	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
nampriknum	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
net	I-Indicator
svc	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
somtum	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
svcws	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
somtum	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
www	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
somtum	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
somtum	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
shop	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
databit	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
svc	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
databit	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
test	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
databit	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
www	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
databit	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
admin	B-Indicator
[	I-Indicator
.databit	I-Indicator
[	I-Indicator
.today	I-Indicator
cendata	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
svc	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
cendata	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
svcws	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
cendata	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
www	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
cendata	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
today	I-Indicator
PHA	O
Family	O
Highlights	O
:	O
Zen	B-Malware
and	O
its	O
cousins	O
January	O
11	O
,	O
2019	O
Google	B-System
Play	I-System
Protect	I-System
detects	O
Potentially	O
Harmful	O
Applications	O
(	O
PHAs	O
)	O
which	O
Google	B-System
Play	I-System
Protect	I-System
defines	O
as	O
any	O
mobile	O
app	O
that	O
poses	O
a	O
potential	O
security	O
risk	O
to	O
users	O
or	O
to	O
user	O
data—commonly	O
referred	O
to	O
as	O
"	O
malware	O
.	O

''	O
in	O
a	O
variety	O
of	O
ways	O
,	O
such	O
as	O
static	O
analysis	O
,	O
dynamic	O
analysis	O
,	O
and	O
machine	O
learning	O
.	O

While	O
our	O
systems	O
are	O
great	O
at	O
automatically	O
detecting	O
and	O
protecting	O
against	O
PHAs	O
,	O
we	O
believe	O
the	O
best	O
security	O
comes	O
from	O
the	O
combination	O
of	O
automated	O
scanning	O
and	O
skilled	O
human	O
review	O
.	O

With	O
this	O
blog	O
series	O
we	O
will	O
be	O
sharing	O
our	O
research	O
analysis	O
with	O
the	O
research	O
and	O
broader	O
security	O
community	O
,	O
starting	O
with	O
the	O
PHA	O
family	O
,	O
Zen	B-Malware
.	O

Zen	B-Malware
uses	O
root	O
permissions	O
on	O
a	O
device	O
to	O
automatically	O
enable	O
a	O
service	O
that	O
creates	O
fake	O
Google	B-Organization
accounts	O
.	O

These	O
accounts	O
are	O
created	O
by	O
abusing	O
accessibility	O
services	O
.	O

Zen	B-Malware
apps	O
gain	O
access	O
to	O
root	O
permissions	O
from	O
a	O
rooting	O
trojan	O
in	O
its	O
infection	O
chain	O
.	O

In	O
this	O
blog	O
post	O
,	O
we	O
do	O
not	O
differentiate	O
between	O
the	O
rooting	O
component	O
and	O
the	O
component	O
that	O
abuses	O
root	O
:	O
we	O
refer	O
to	O
them	O
interchangeably	O
as	O
Zen	B-Malware
.	O

We	O
also	O
describe	O
apps	O
that	O
we	O
think	O
are	O
coming	O
from	O
the	O
same	O
author	O
or	O
a	O
group	O
of	O
authors	O
.	O

All	O
of	O
the	O
PHAs	O
that	O
are	O
mentioned	O
in	O
this	O
blog	O
post	O
were	O
detected	O
and	O
removed	O
by	O
Google	B-System
Play	I-System
Protect	I-System
.	O

Background	O
Uncovering	O
PHAs	O
takes	O
a	O
lot	O
of	O
detective	O
work	O
and	O
unraveling	O
the	O
mystery	O
of	O
how	O
they	O
're	O
possibly	O
connected	O
to	O
other	O
apps	O
takes	O
even	O
more	O
.	O

PHA	O
authors	O
usually	O
try	O
to	O
hide	O
their	O
tracks	O
,	O
so	O
attribution	O
is	O
difficult	O
.	O

Sometimes	O
,	O
we	O
can	O
attribute	O
different	O
apps	O
to	O
the	O
same	O
author	O
based	O
on	O
a	O
small	O
,	O
unique	O
pieces	O
of	O
evidence	O
that	O
suggest	O
similarity	O
,	O
such	O
as	O
a	O
repetition	O
of	O
an	O
exceptionally	O
rare	O
code	O
snippet	O
,	O
asset	O
,	O
or	O
a	O
particular	O
string	O
in	O
the	O
debug	O
logs	O
.	O

Every	O
once	O
in	O
a	O
while	O
,	O
authors	O
leave	O
behind	O
a	O
trace	O
that	O
allows	O
us	O
to	O
attribute	O
not	O
only	O
similar	O
apps	O
,	O
but	O
also	O
multiple	O
different	O
PHA	O
families	O
to	O
the	O
same	O
group	O
or	O
person	O
.	O

However	O
,	O
the	O
actual	O
timeline	O
of	O
the	O
creation	O
of	O
different	O
variants	O
is	O
unclear	O
.	O

In	O
April	O
2013	O
,	O
we	O
saw	O
the	O
first	O
sample	O
,	O
which	O
made	O
heavy	O
use	O
of	O
dynamic	O
code	O
loading	O
(	O
i.e.	O
,	O
fetching	O
executable	O
code	O
from	O
remote	O
sources	O
after	O
the	O
initial	O
app	O
is	O
installed	O
)	O
.	O

Dynamic	O
code	O
loading	O
makes	O
it	O
impossible	O
to	O
state	O
what	O
kind	O
of	O
PHA	O
it	O
was	O
.	O

This	O
sample	O
displayed	O
ads	O
from	O
various	O
sources	O
.	O

More	O
recent	O
variants	O
blend	O
rooting	O
capabilities	O
and	O
click	O
fraud	O
.	O

As	O
rooting	O
exploits	O
on	O
Android	B-System
become	O
less	O
prevalent	O
and	O
lucrative	O
,	O
PHA	O
authors	O
adapt	O
their	O
abuse	O
or	O
monetization	O
strategy	O
to	O
focus	O
on	O
tactics	O
like	O
click	O
fraud	O
.	O

This	O
post	O
does	O
n't	O
follow	O
the	O
chronological	O
evolution	O
of	O
Zen	B-Malware
,	O
but	O
instead	O
covers	O
relevant	O
samples	O
from	O
least	O
to	O
most	O
complex	O
.	O

Apps	O
with	O
a	O
custom-made	O
advertisement	O
SDK	O
The	O
simplest	O
PHA	O
from	O
the	O
author	O
's	O
portfolio	O
used	O
a	O
specially	O
crafted	O
advertisement	O
SDK	O
to	O
create	O
a	O
proxy	O
for	O
all	O
ads-related	O
network	O
traffic	O
.	O

By	O
proxying	O
all	O
requests	O
through	O
a	O
custom	O
server	O
,	O
the	O
real	O
source	O
of	O
ads	O
is	O
opaque	O
.	O

This	O
example	O
shows	O
one	O
possible	O
implementation	O
of	O
this	O
technique	O
.	O

This	O
approach	O
allows	O
the	O
authors	O
to	O
combine	O
ads	O
from	O
third-party	O
advertising	O
networks	O
with	O
ads	O
they	O
created	O
for	O
their	O
own	O
apps	O
.	O

It	O
may	O
even	O
allow	O
them	O
to	O
sell	O
ad	O
space	O
directly	O
to	O
application	O
developers	O
.	O

The	O
advertisement	O
SDK	O
also	O
collects	O
statistics	O
about	O
clicks	O
and	O
impressions	O
to	O
make	O
it	O
easier	O
to	O
track	O
revenue	O
.	O

Selling	O
the	O
ad	O
traffic	O
directly	O
or	O
displaying	O
ads	O
from	O
other	O
sources	O
in	O
a	O
very	O
large	O
volume	O
can	O
provide	O
direct	O
profit	O
to	O
the	O
app	O
author	O
from	O
the	O
advertisers	O
.	O

We	O
have	O
seen	O
two	O
types	O
of	O
apps	O
that	O
use	O
this	O
custom-made	O
SDK	O
.	O

The	O
first	O
are	O
games	O
of	O
very	O
low	O
quality	O
that	O
mimic	O
the	O
experience	O
of	O
popular	O
mobile	O
games	O
.	O

While	O
the	O
counterfeit	O
games	O
claim	O
to	O
provide	O
similar	O
functionality	O
to	O
the	O
popular	O
apps	O
,	O
they	O
are	O
simply	O
used	O
to	O
display	O
ads	O
through	O
a	O
custom	O
advertisement	O
SDK	O
.	O

The	O
second	O
type	O
of	O
apps	O
reveals	O
an	O
evolution	O
in	O
the	O
author	O
's	O
tactics	O
.	O

Instead	O
of	O
implementing	O
very	O
basic	O
gameplay	O
,	O
the	O
authors	O
pirated	O
and	O
repackaged	O
the	O
original	O
game	O
in	O
their	O
app	O
and	O
bundled	O
with	O
it	O
their	O
advertisement	O
SDK	O
.	O

The	O
only	O
noticeable	O
difference	O
is	O
the	O
game	O
has	O
more	O
ads	O
,	O
including	O
ads	O
on	O
the	O
very	O
first	O
screen	O
.	O

In	O
all	O
cases	O
,	O
the	O
ads	O
are	O
used	O
to	O
convince	O
users	O
to	O
install	O
other	O
apps	O
from	O
different	O
developer	O
accounts	O
,	O
but	O
written	O
by	O
the	O
same	O
group	O
.	O

Those	O
apps	O
use	O
the	O
same	O
techniques	O
to	O
monetize	O
their	O
actions	O
.	O

Click	O
fraud	O
apps	O
The	O
authors	O
'	O
tactics	O
evolved	O
from	O
advertisement	O
spam	O
to	O
real	O
PHA	O
(	O
Click	O
Fraud	O
)	O
.	O

Click	O
fraud	O
PHAs	O
simulate	O
user	O
clicks	O
on	O
ads	O
instead	O
of	O
simply	O
displaying	O
ads	O
and	O
waiting	O
for	O
users	O
to	O
click	O
them	O
.	O

This	O
allows	O
the	O
PHA	O
authors	O
to	O
monetize	O
their	O
apps	O
more	O
effectively	O
than	O
through	O
regular	O
advertising	O
.	O

This	O
behavior	O
negatively	O
impacts	O
advertisement	O
networks	O
and	O
their	O
clients	O
because	O
advertising	O
budget	O
is	O
spent	O
without	O
acquiring	O
real	O
customers	O
,	O
and	O
impacts	O
user	O
experience	O
by	O
consuming	O
their	O
data	O
plan	O
resources	O
.	O

The	O
click	O
fraud	O
PHA	O
requests	O
a	O
URL	O
to	O
the	O
advertising	O
network	O
directly	O
instead	O
of	O
proxying	O
it	O
through	O
an	O
additional	O
SDK	O
.	O

The	O
command	O
&	O
control	O
server	O
(	O
C	O
&	O
C	O
server	O
)	O
returns	O
the	O
URL	O
to	O
click	O
along	O
with	O
a	O
very	O
long	O
list	O
of	O
additional	O
parameters	O
in	O
JSON	O
format	O
.	O

After	O
rendering	O
the	O
ad	O
on	O
the	O
screen	O
,	O
the	O
app	O
tries	O
to	O
identify	O
the	O
part	O
of	O
the	O
advertisement	O
website	O
to	O
click	O
.	O

If	O
that	O
part	O
is	O
found	O
,	O
the	O
app	O
loads	O
Javascript	O
snippets	O
from	O
the	O
JSON	O
parameters	O
to	O
click	O
a	O
button	O
or	O
other	O
HTML	O
element	O
,	O
simulating	O
a	O
real	O
user	O
click	O
.	O

Because	O
a	O
user	O
interacting	O
with	O
an	O
ad	O
often	O
leads	O
to	O
a	O
higher	O
chance	O
of	O
the	O
user	O
purchasing	O
something	O
,	O
ad	O
networks	O
often	O
"	O
pay	O
per	O
click	O
''	O
to	O
developers	O
who	O
host	O
their	O
ads	O
.	O

Therefore	O
,	O
by	O
simulating	O
fraudulent	O
clicks	O
,	O
these	O
developers	O
are	O
making	O
money	O
without	O
requiring	O
a	O
user	O
to	O
click	O
on	O
an	O
advertisement	O
.	O

This	O
example	O
code	O
shows	O
a	O
JSON	O
reply	O
returned	O
by	O
the	O
C	O
&	O
C	O
server	O
.	O

It	O
has	O
been	O
shortened	O
for	O
brevity	O
.	O

Based	O
on	O
this	O
JSON	O
reply	O
,	O
the	O
app	O
looks	O
for	O
an	O
HTML	O
snippet	O
that	O
corresponds	O
to	O
the	O
active	O
element	O
(	O
show_hide	O
btnnext	O
)	O
and	O
,	O
if	O
found	O
,	O
the	O
Javascript	O
snippet	O
tries	O
to	O
perform	O
a	O
click	O
(	O
)	O
method	O
on	O
it	O
.	O

Rooting	O
trojans	O
The	O
Zen	B-Malware
authors	O
have	O
also	O
created	O
a	O
rooting	O
trojan	O
.	O

Using	O
a	O
publicly	O
available	O
rooting	O
framework	O
,	O
the	O
PHA	O
attempts	O
to	O
root	O
devices	O
and	O
gain	O
persistence	O
on	O
them	O
by	O
reinstalling	O
itself	O
on	O
the	O
system	O
partition	O
of	O
rooted	O
device	O
.	O

Installing	O
apps	O
on	O
the	O
system	O
partition	O
makes	O
it	O
harder	O
for	O
the	O
user	O
to	O
remove	O
the	O
app	O
.	O

This	O
technique	O
only	O
works	O
for	O
unpatched	O
devices	O
running	O
Android	B-System
4.3	I-System
or	O
lower	O
.	O

Devices	O
running	O
Android	B-System
4.4	I-System
and	O
higher	O
are	O
protected	O
by	O
Verified	O
Boot	O
.	O

Zen	B-Malware
's	O
rooting	O
trojan	O
apps	O
target	O
a	O
specific	O
device	O
model	O
with	O
a	O
very	O
specific	O
system	O
image	O
.	O

After	O
achieving	O
root	O
access	O
the	O
app	O
tries	O
to	O
replace	O
the	O
framework.jar	B-Indicator
file	O
on	O
the	O
system	O
partition	O
.	O

Replicating	O
framework.jar	B-Indicator
allows	O
the	O
app	O
to	O
intercept	O
and	O
modify	O
the	O
behavior	O
of	O
the	O
Android	B-System
standard	O
API	O
.	O

In	O
particular	O
,	O
these	O
apps	O
try	O
to	O
add	O
an	O
additional	O
method	O
called	O
statistics	O
(	O
)	O
into	O
the	O
Activity	O
class	O
.	O

When	O
inserted	O
,	O
this	O
method	O
runs	O
every	O
time	O
any	O
Activity	O
object	O
in	O
any	O
Android	O
app	O
is	O
created	O
.	O

This	O
happens	O
all	O
the	O
time	O
in	O
regular	O
Android	B-System
apps	O
,	O
as	O
Activity	O
is	O
one	O
of	O
the	O
fundamental	O
Android	B-System
UI	O
elements	O
.	O

The	O
only	O
purpose	O
of	O
this	O
method	O
is	O
to	O
connect	O
to	O
the	O
C	O
&	O
C	O
server	O
.	O

The	O
Zen	B-Malware
trojan	O
After	O
achieving	O
persistence	O
,	O
the	O
trojan	O
downloads	O
additional	O
payloads	O
,	O
including	O
another	O
trojan	O
called	O
Zen	B-Malware
.	O

Zen	B-Malware
requires	O
root	O
to	O
work	O
correctly	O
on	O
the	O
Android	B-System
operating	O
system	O
.	O

The	O
Zen	B-Malware
trojan	O
uses	O
its	O
root	O
privileges	O
to	O
turn	O
on	O
accessibility	O
service	O
(	O
a	O
service	O
used	O
to	O
allow	O
Android	B-System
users	O
with	O
disabilities	O
to	O
use	O
their	O
devices	O
)	O
for	O
itself	O
by	O
writing	O
to	O
a	O
system-wide	O
setting	O
value	O
enabled_accessibility_services	O
.	O

Zen	B-Malware
does	O
n't	O
even	O
check	O
for	O
the	O
root	O
privilege	O
:	O
it	O
just	O
assumes	O
it	O
has	O
it	O
.	O

This	O
leads	O
us	O
to	O
believe	O
that	O
Zen	B-Malware
is	O
just	O
part	O
of	O
a	O
larger	O
infection	O
chain	O
.	O

The	O
trojan	O
implements	O
three	O
accessibility	O
services	O
directed	O
at	O
different	O
Android	B-System
API	I-System
levels	O
and	O
uses	O
these	O
accessibility	O
services	O
,	O
chosen	O
by	O
checking	O
the	O
operating	O
system	O
version	O
,	O
to	O
create	O
new	O
Google	B-Organization
accounts	O
.	O

This	O
is	O
done	O
by	O
opening	O
the	O
Google	B-Organization
account	O
creation	O
process	O
and	O
parsing	O
the	O
current	O
view	O
.	O

The	O
app	O
then	O
clicks	O
the	O
appropriate	O
buttons	O
,	O
scrollbars	O
,	O
and	O
other	O
UI	O
elements	O
to	O
go	O
through	O
account	O
sign-up	O
without	O
user	O
intervention	O
.	O

During	O
the	O
account	O
sign-up	O
process	O
,	O
Google	B-Organization
may	O
flag	O
the	O
account	O
creation	O
attempt	O
as	O
suspicious	O
and	O
prompt	O
the	O
app	O
to	O
solve	O
a	O
CAPTCHA	O
.	O

To	O
get	O
around	O
this	O
,	O
the	O
app	O
then	O
uses	O
its	O
root	O
privilege	O
to	O
inject	O
code	O
into	O
the	O
Setup	O
Wizard	O
,	O
extract	O
the	O
CAPTCHA	O
image	O
,	O
and	O
sends	O
it	O
to	O
a	O
remote	O
server	O
to	O
try	O
to	O
solve	O
the	O
CAPTCHA	O
.	O

It	O
is	O
unclear	O
if	O
the	O
remote	O
server	O
is	O
capable	O
of	O
solving	O
the	O
CAPTCHA	O
image	O
automatically	O
or	O
if	O
this	O
is	O
done	O
manually	O
by	O
a	O
human	O
in	O
the	O
background	O
.	O

After	O
the	O
server	O
returns	O
the	O
solution	O
,	O
the	O
app	O
enters	O
it	O
into	O
the	O
appropriate	O
text	O
field	O
to	O
complete	O
the	O
CAPTCHA	O
challenge	O
.	O

The	O
Zen	B-Malware
trojan	O
does	O
not	O
implement	O
any	O
kind	O
of	O
obfuscation	O
except	O
for	O
one	O
string	O
that	O
is	O
encoded	O
using	O
Base64	O
encoding	O
.	O

It	O
's	O
one	O
of	O
the	O
strings	O
-	O
"	O
How	O
you	O
'll	O
sign	O
in	O
''	O
-	O
that	O
it	O
looks	O
for	O
during	O
the	O
account	O
creation	O
process	O
.	O

The	O
code	O
snippet	O
below	O
shows	O
part	O
of	O
the	O
screen	O
parsing	O
process	O
.	O

Apart	O
from	O
injecting	O
code	O
to	O
read	O
the	O
CAPTCHA	O
,	O
the	O
app	O
also	O
injects	O
its	O
own	O
code	O
into	O
the	O
system_server	O
process	O
,	O
which	O
requires	O
root	O
privileges	O
.	O

This	O
indicates	O
that	O
the	O
app	O
tries	O
to	O
hide	O
itself	O
from	O
any	O
anti-PHA	O
systems	O
that	O
look	O
for	O
a	O
specific	O
app	O
process	O
name	O
or	O
does	O
not	O
have	O
the	O
ability	O
to	O
scan	O
the	O
memory	O
of	O
the	O
system_server	O
process	O
.	O

The	O
app	O
also	O
creates	O
hooks	O
to	O
prevent	O
the	O
phone	O
from	O
rebooting	O
,	O
going	O
to	O
sleep	O
or	O
allowing	O
the	O
user	O
from	O
pressing	O
hardware	O
buttons	O
during	O
the	O
account	O
creation	O
process	O
.	O

These	O
hooks	O
are	O
created	O
using	O
the	O
root	O
access	O
and	O
a	O
custom	O
native	O
code	O
called	O
Lmt_INJECT	O
,	O
although	O
the	O
algorithm	O
for	O
this	O
is	O
well	O
known	O
.	O

First	O
,	O
the	O
app	O
has	O
to	O
turn	O
off	O
SELinux	B-System
protection	O
.	O

Then	O
the	O
app	O
finds	O
a	O
process	O
id	O
value	O
for	O
the	O
process	O
it	O
wants	O
to	O
inject	O
with	O
code	O
.	O

This	O
is	O
done	O
using	O
a	O
series	O
of	O
syscalls	O
as	O
outlined	O
below	O
.	O

The	O
"	O
source	O
process	O
''	O
refers	O
to	O
the	O
Zen	B-Malware
trojan	O
running	O
as	O
root	O
,	O
while	O
the	O
"	O
target	O
process	O
''	O
refers	O
to	O
the	O
process	O
to	O
which	O
the	O
code	O
is	O
injected	O
and	O
[	O
pid	O
]	O
refers	O
to	O
the	O
target	O
process	O
pid	O
value	O
.	O

The	O
source	O
process	O
checks	O
the	O
mapping	O
between	O
a	O
process	O
id	O
and	O
a	O
process	O
name	O
.	O

This	O
is	O
done	O
by	O
reading	O
the	O
/proc/	B-Indicator
[	I-Indicator
pid	I-Indicator
]	I-Indicator
/cmdline	I-Indicator
file	O
.	O

This	O
very	O
first	O
step	O
fails	O
in	O
Android	B-System
7.0	I-System
and	O
higher	O
,	O
even	O
with	O
a	O
root	O
permission	O
.	O

The	O
/proc	B-Indicator
filesystem	O
is	O
now	O
mounted	O
with	O
a	O
hidepid=2	O
parameter	O
,	O
which	O
means	O
that	O
the	O
process	O
can	O
not	O
access	O
other	O
process	O
/proc/	B-Indicator
[	I-Indicator
pid	I-Indicator
]	I-Indicator
directory	O
.	O

A	O
ptrace_attach	O
syscall	O
is	O
called	O
.	O

This	O
allows	O
the	O
source	O
process	O
to	O
trace	O
the	O
target	O
.	O

The	O
source	O
process	O
looks	O
at	O
its	O
own	O
memory	O
to	O
calculate	O
the	O
offset	O
between	O
the	O
beginning	O
of	O
the	O
libc	O
library	O
and	O
the	O
mmap	O
address	O
.	O

The	O
source	O
process	O
reads	O
/proc/	B-Indicator
[	I-Indicator
pid	I-Indicator
]	I-Indicator
/maps	I-Indicator
to	O
find	O
where	O
libc	O
is	O
located	O
in	O
the	O
target	O
process	O
memory	O
.	O

By	O
adding	O
the	O
previously	O
calculated	O
offset	O
,	O
it	O
can	O
get	O
the	O
address	O
of	O
the	O
mmap	O
function	O
in	O
the	O
target	O
process	O
memory	O
.	O

The	O
source	O
process	O
tries	O
to	O
determine	O
the	O
location	O
of	O
dlopen	O
,	O
dlsym	O
,	O
and	O
dlclose	O
functions	O
in	O
the	O
target	O
process	O
.	O

It	O
uses	O
the	O
same	O
technique	O
as	O
it	O
used	O
to	O
determine	O
the	O
offset	O
to	O
the	O
mmap	O
function	O
.	O

The	O
source	O
process	O
writes	O
the	O
native	O
shellcode	O
into	O
the	O
memory	O
region	O
allocated	O
by	O
mmap	O
.	O

Additionally	O
,	O
it	O
also	O
writes	O
addresses	O
of	O
dlopen	O
,	O
dlsym	O
,	O
and	O
dlclose	O
into	O
the	O
same	O
region	O
,	O
so	O
that	O
they	O
can	O
be	O
used	O
by	O
the	O
shellcode	O
.	O

Shellcode	O
simply	O
uses	O
dlopen	O
to	O
open	O
a	O
.so	O
file	O
within	O
the	O
target	O
process	O
and	O
then	O
dlsym	O
to	O
find	O
a	O
symbol	B-Organization
in	O
that	O
file	O
and	O
run	O
it	O
.	O

The	O
source	O
process	O
changes	O
the	O
registers	O
in	O
the	O
target	O
process	O
so	O
that	O
PC	O
register	O
points	O
directly	O
to	O
the	O
shellcode	O
.	O

This	O
is	O
done	O
using	O
the	O
ptrace	O
syscall	O
.	O

This	O
diagram	O
illustrates	O
the	O
whole	O
process	O
.	O

Summary	O
PHA	O
authors	O
go	O
to	O
great	O
lengths	O
to	O
come	O
up	O
with	O
increasingly	O
clever	O
ways	O
to	O
monetize	O
their	O
apps	O
.	O

Zen	B-Malware
family	O
PHA	O
authors	O
exhibit	O
a	O
wide	O
range	O
of	O
techniques	O
,	O
from	O
simply	O
inserting	O
an	O
advertising	O
SDK	O
to	O
a	O
sophisticated	O
trojan	O
.	O

The	O
app	O
that	O
resulted	O
in	O
the	O
largest	O
number	O
of	O
affected	O
users	O
was	O
the	O
click	O
fraud	O
version	O
,	O
which	O
was	O
installed	O
over	O
170,000	O
times	O
at	O
its	O
peak	O
in	O
February	O
2018	O
.	O

The	O
most	O
affected	O
countries	O
were	O
India	O
,	O
Brazil	O
,	O
and	O
Indonesia	O
.	O

In	O
most	O
cases	O
,	O
these	O
click	O
fraud	O
apps	O
were	O
uninstalled	O
by	O
the	O
users	O
,	O
probably	O
due	O
to	O
the	O
low	O
quality	O
of	O
the	O
apps	O
.	O

If	O
Google	B-System
Play	I-System
Protect	I-System
detects	O
one	O
of	O
these	O
apps	O
,	O
Google	B-System
Play	I-System
Protect	I-System
will	O
show	O
a	O
warning	O
to	O
users	O
.	O

We	O
are	O
constantly	O
on	O
the	O
lookout	O
for	O
new	O
threats	O
and	O
we	O
are	O
expanding	O
our	O
protections	O
.	O

Every	O
device	O
with	O
Google	B-System
Play	I-System
includes	O
Google	B-System
Play	I-System
Protect	I-System
and	O
all	O
apps	O
on	O
Google	B-System
Play	I-System
are	O
automatically	O
and	O
periodically	O
scanned	O
by	O
our	O
solutions	O
.	O

You	O
can	O
check	O
the	O
status	O
of	O
Google	B-System
Play	I-System
Protect	I-System
on	O
your	O
device	O
:	O
Open	O
your	O
Android	O
device	O
's	O
Google	B-System
Play	I-System
Store	I-System
app	O
.	O

Tap	O
Menu	O
>	O
Play	O
Protect	O
.	O

Look	O
for	O
information	O
about	O
the	O
status	O
of	O
your	O
device	O
.	O

Hashes	O
of	O
samples	O
Type	O
Package	O
name	O
SHA256	O
digest	O
Custom	O
ads	O
com.targetshoot.zombieapocalypse.sniper.zombieshootinggame	B-Indicator
5d98d8a7a012a858f0fa4cf8d2ed3d5a82937b1a98ea2703d440307c63c6c928	B-Indicator
Click	O
fraud	O
com.counterterrorist.cs.elite.combat.shootinggame	B-Indicator
84672fb2f228ec749d3c3c1cb168a1c31f544970fd29136bea2a5b2cefac6d04	B-Indicator

Rooting	O
trojan	O
com.android.world.news	O
bd233c1f5c477b0cc15d7f84392dab3a7a598243efa3154304327ff4580ae213	B-Indicator
Zen	B-Malware
trojan	O
com.lmt.register	B-Indicator
eb12cd65589cbc6f9d3563576c304273cb6a78072b0c20a155a0951370476d8d	B-Indicator
Mobile	O
Campaign	O
‘	O
Bouncing	B-Malware
Golf	I-Malware
’	O
Affects	O
Middle	O
East	O
We	O
uncovered	O
a	O
cyberespionage	O
campaign	O
targeting	O
Middle	O

Eastern	O
countries	O
.	O

We	O
named	O
this	O
campaign	O
“	O
Bouncing	B-Malware
Golf	I-Malware
”	O
based	O
on	O
the	O
malware	O
’	O
s	O
code	O
in	O
the	O
package	O
named	O
“	O
golf.	O
”	O
June	O
18	O
,	O
2019	O
We	O
uncovered	O
a	O
cyberespionage	O
campaign	O
targeting	O
Middle	O
Eastern	O
countries	O
.	O

We	O
named	O
this	O
campaign	O
“	O
Bouncing	B-Malware
Golf	I-Malware
”	O
based	O
on	O
the	O
malware	O
’	O
s	O
code	O
in	O
the	O
package	O
named	O
“	O
golf.	O
”	O
The	O
malware	O
involved	O
,	O
which	O
Trend	B-Organization
Micro	I-Organization
detects	O
as	O
AndroidOS_GolfSpy.HRX	B-Malware
,	O
is	O
notable	O
for	O
its	O
wide	O
range	O
of	O
cyberespionage	O
capabilities	O
.	O

Malicious	O
codes	O
are	O
embedded	O
in	O
apps	O
that	O
the	O
operators	O
repackaged	O
from	O
legitimate	O
applications	O
.	O

Monitoring	O
the	O
command	O
and	O
control	O
(	O
C	O
&	O
C	O
)	O
servers	O
used	O
by	O
Bouncing	B-Malware
Golf	I-Malware
,	O
we	O
’	O
ve	O
so	O
far	O
observed	O
more	O
than	O
660	O
Android	B-System
devices	O
infected	O
with	O
GolfSpy	B-Malware
.	O

Much	O
of	O
the	O
information	O
being	O
stolen	O
appear	O
to	O
be	O
military-related	O
.	O

The	O
campaign	O
’	O
s	O
attack	O
vector	O
is	O
also	O
interesting	O
.	O

These	O
repackaged	O
,	O
malware-laden	O
apps	O
are	O
neither	O
on	O
Google	B-System
Play	I-System
nor	O
popular	O
third-party	O
app	O
marketplaces	O
,	O
and	O
we	O
only	O
saw	O
the	O
website	O
hosting	O
the	O
malicious	O
apps	O
being	O
promoted	O
on	O
social	O
media	O
when	O
we	O
followed	O
GolfSpy	B-Malware
’	O
s	O
trail	O
.	O

We	O
were	O
also	O
able	O
to	O
analyze	O
some	O
GolfSpy	B-Malware
samples	O
sourced	O
from	O
the	O
Trend	B-Organization
Micro	I-Organization
mobile	O
app	O
reputation	O
service	O
.	O

Also	O
of	O
note	O
is	O
Bouncing	B-Malware
Golf	I-Malware
’	O
s	O
possible	O
connection	O
to	O
a	O
previously	O
reported	O
mobile	O
cyberespionage	O
campaign	O
that	O
researchers	O
named	O
Domestic	B-Malware
Kitten	I-Malware
.	O

The	O
strings	O
of	O
code	O
,	O
for	O
one	O
,	O
are	O
similarly	O
structured	O
.	O

The	O
data	O
targeted	O
for	O
theft	O
also	O
have	O
similar	O
formats	O
.	O

Figure	O
1	O
.	O

GolfSpy	B-Malware
’	O
s	O
infection	O
chain	O
GolfSpy	B-Malware
's	O
Potential	O
Impact	O
Given	O
GolfSpy	B-Malware
’	O
s	O
information-stealing	O
capabilities	O
,	O
this	O
malware	O
can	O
effectively	O
hijack	O
an	O
infected	O
Android	B-System
device	O
.	O

Here	O
is	O
a	O
list	O
of	O
information	O
that	O
GolfSpy	B-Malware
steals	O
:	O
Device	O
accounts	O
List	O
of	O
applications	O
installed	O
in	O
the	O
device	O
Device	O
’	O
s	O
current	O
running	O
processes	O
Battery	O
status	O
Bookmarks/Histories	O
of	O
the	O
device	O
’	O
s	O
default	O
browser	O
Call	O
logs	O
and	O
records	O
Clipboard	O
contents	O
Contacts	O
,	O
including	O
those	O
in	O
VCard	O
format	O
Mobile	O
operator	O
information	O
Files	O
stored	O
on	O
SDcard	O
Device	O
location	O
List	O
of	O
image	O
,	O
audio	O
,	O
and	O
video	O
files	O
stored	O
on	O
the	O
device	O
Storage	O
and	O
memory	O
information	O
Connection	O
information	O
Sensor	O
information	O
SMS	O
messages	O
Pictures	O
GolfSpy	B-Malware
also	O
has	O
a	O
function	O
that	O
lets	O
it	O
connect	O
to	O
a	O
remote	O
server	O
to	O
fetch	O
and	O
perform	O
commands	O

,	O
including	O
:	O
searching	O
for	O
,	O
listing	O
,	O
deleting	O
,	O
and	O
renaming	O
files	O
as	O
well	O
as	O
downloading	O
a	O
file	O
into	O
and	O
retrieving	O
a	O
file	O
from	O
the	O
device	O
;	O
taking	O
screenshots	O
;	O
installing	O
other	O
application	O
packages	O
(	O
APK	O
)	O
;	O
recording	O
audio	O
and	O
video	O
;	O
and	O
updating	O
the	O
malware	O
.	O

Technical	O
Analysis	O
The	O
repackaged	O
applications	O
are	O
embedded	O
with	O
malicious	O
code	O
,	O
which	O
can	O
be	O
found	O
in	O
the	O
com.golf	B-Indicator
package	O
.	O

These	O
repackaged	O
apps	O
pose	O
as	O
communication	O
,	O
news	O
,	O
lifestyle	O
,	O
book	O
,	O
and	O
reference	O
apps	O
popularly	O
used	O
in	O
the	O
Middle	O
East	O
.	O

The	O
GolfSpy	B-Malware
malware	O
embedded	O
in	O
the	O
apps	O
is	O
hardcoded	O
with	O
an	O
internal	O
name	O
used	O
by	O
the	O
attacker	O
.	O

Figure	O
2	O
.	O

Icons	O
of	O
the	O
apps	O
that	O
Bouncing	B-Malware
Golf	I-Malware
’	O
s	O
operators	O
repackaged	O
(	O
top	O
)	O
and	O
a	O
comparison	O
of	O
packages	O
between	O
the	O
original	O
legitimate	O
app	O
(	O
bottom	O
left	O
)	O
and	O
GolfSpy	B-Malware
(	O
bottom	O
right	O
)	O
Figure	O
3	O
.	O

GolfSpy	B-Malware
’	O
s	O
configurations	O
encoded	O
by	O
a	O
custom	O
algorithm	O
(	O
right	O
)	O
and	O
its	O
decoded	O
version	O
(	O
left	O
)	O
As	O
shown	O
in	O
Figure	O
3	O
,	O
GolfSpy	B-Malware
’	O
s	O
configurations	O
(	O
e.g.	O
,	O
C	O
&	O
C	O
server	O
,	O
secret	O
keys	O
)	O
are	O
encoded	O
by	O
a	O
customized	O
algorithm	O
.	O

After	O
it	O
is	O
launched	O
,	O
GolfSpy	B-Malware
will	O
generate	O
a	O
unique	O
ID	O
for	O
the	O
affected	O
device	O
and	O
then	O
collect	O
its	O
data	O
such	O
as	O
SMS	O
,	O
contact	O
list	O
,	O
location	O
,	O
and	O
accounts	O
in	O
this	O
format	O
:	O
“	O
%	O
,	O
[	O
]	O
,	O
time	O
”	O
(	O
shown	O
in	O
Figure	O
4	O
)	O
.	O

The	O
information	O
is	O
written	O
into	O
a	O
file	O
on	O
the	O
device	O
.	O

The	O
attacker	O
can	O
choose	O
the	O
data	O
types	O
to	O
collect	O
,	O
which	O
are	O
written	O
in	O
a	O
certain	O
format	O
.	O

Figure	O
4	O
.	O

Code	O
snippet	O
showing	O
GolfSpy	B-Malware
generating	O
UUID	O
The	O
value	O
of	O
%	O
is	O
in	O
the	O
range	O
of	O
1-9	O
or	O
a-j	O
.	O

Each	O
value	O
represents	O
a	O
different	O
type	O
of	O
data	O
to	O
steal	O
from	O
the	O
device	O
:	O
Value	O
Data	O
Type	O
1	O
Accounts	O
2	O
Installed	O
APP	O
list	O
3	O
Running	O
processes	O
list	O
4	O
Battery	O
status	O
5	O
Browser	O
bookmarks	O
and	O
histories	O
6	O
Call	O
logs	O
7	O
Clipboard	O
8	O
Contacts	O
9	O
Mobile	O
operator	O
information	O
a	O
File	O
list	O
on	O
SD	O
card	O
b	O
Location	O
c	O
Image	O
list	O
d	O
Audio	O
list	O
e	O
Video	O
list	O
f	O
Storage	O
and	O
memory	O
information	O
g	O
Connection	O
information	O
h	O
Sensors	O
information	O
i	O
SMS	O
messages	O
j	O
VCard	O
format	O
contacts	O
Table	O
1	O
.	O

The	O
type	O
of	O
data	O
corresponding	O
to	O
the	O
value	O
coded	O
in	O
GolfSpy	B-Malware
Figure	O
5	O
shows	O
the	O
code	O
snippets	O
that	O
are	O
involved	O
in	O
monitoring	O
and	O
recording	O
the	O
device	O
’	O
s	O
phone	O
call	O
.	O

It	O
will	O
also	O
take	O
a	O
photo	O
using	O
the	O
device	O
’	O
s	O
front	O
camera	O
when	O
the	O
user	O
wakes	O
the	O
device	O
.	O

Apart	O
from	O
collecting	O
the	O
above	O
data	O
,	O
the	O
spyware	O
monitors	O
users	O
’	O
phone	O
calls	O
,	O
records	O
them	O
,	O
and	O
saves	O
the	O
recorded	O
file	O
on	O
the	O
device	O
.	O

GolfSpy	B-Malware
encrypts	O
all	O
the	O
stolen	O
data	O
using	O
a	O
simple	O
XOR	O
operation	O
with	O
a	O
pre-configured	O
key	O
before	O
sending	O
it	O
to	O
the	O
C	O
&	O
C	O
server	O
using	O
the	O
HTTP	O
POST	O
method	O
.	O

Figure	O
5	O
.	O

Code	O
snippets	O
showing	O
how	O
GolfSpy	B-Malware
monitors	O
phone	O
calls	O
via	O
register	O
receiver	O
(	O
top	O
left	O
)	O
,	O
its	O
actions	O
when	O
the	O
device	O
is	O
woken	O
up	O
(	O
top	O
right	O
)	O
,	O
and	O
how	O
it	O
encrypts	O
the	O
stolen	O
data	O
(	O
bottom	O
)	O
The	O
malware	O
retrieves	O
commands	O
from	O
the	O
C	O
&	O
C	O
server	O
via	O
HTTP	O
,	O
and	O
attackers	O
can	O
steal	O
specific	O
files	O
on	O
the	O
infected	O
device	O
.	O

The	O
command	O
is	O
a	O
constructed	O
string	O
split	O
into	O
three	O
parts	O
using	O
"	O
"	O
as	O
a	O
separator	O
.	O

The	O
first	O
part	O
is	O
the	O
target	O
directory	O
,	O
the	O
second	O
is	O
a	O
regular	O
expression	O
used	O
to	O
match	O
specific	O
files	O
,	O
while	O
the	O
last	O
part	O
is	O
an	O
ID	O
.	O

Figure	O
6	O
.	O

Example	O
of	O
a	O
command	O
that	O
steals	O
specific	O
files	O
from	O
an	O
infected	O
device	O
’	O
s	O
application	O
(	O
top	O
)	O
,	O
and	O
GolfSpy	B-Malware
’	O
s	O
parse-and-perform	O
command	O
(	O
bottom	O
)	O
Apart	O
from	O
the	O
HTTP	O
POST	O
method	O
,	O
GolfSpy	B-Malware
also	O
creates	O
a	O
socket	O
connection	O
to	O
the	O
remote	O
C	O
&	O
C	O
server	O
in	O
order	O
to	O
receive	O
and	O
perform	O
additional	O
commands	O
.	O

Stolen	O
data	O
will	O
also	O
be	O
encrypted	O
and	O
sent	O
to	O
the	O
C	O
&	O
C	O
server	O
via	O
the	O
socket	O
connection	O
.	O

The	O
encryption	O
key	O
is	O
different	O
from	O
the	O
one	O
used	O
for	O
sending	O
stolen	O
data	O
via	O
HTTP	O
.	O

Figure	O
7	O
.	O

The	O
additional	O
commands	O
that	O
attackers	O
can	O
carry	O
out	O
via	O
a	O
socket	O
connection	O
(	O
top	O
)	O
and	O
the	O
key	O
used	O
to	O
encrypt	O
the	O
stolen	O
data	O
(	O
bottom	O
)	O
Correlating	O
Bouncing	B-Malware
Golf	I-Malware
's	O
Activities	O
We	O
monitored	O
Bouncing	B-Malware
Golf	I-Malware
’	O
s	O
C	O
&	O
C-related	O
activities	O
and	O
saw	O
that	O
the	O
campaign	O
has	O
affected	O
more	O
than	O
660	O
devices	O
as	O
of	O
this	O
writing	O
.	O

The	O
small	O
or	O
limited	O
number	O
is	O
understandable	O
given	O
the	O
nature	O
of	O
this	O
campaign	O
,	O
but	O
we	O
also	O
expect	O
it	O
to	O
increase	O
or	O
even	O
diversify	O
in	O
terms	O
of	O
distribution	O
.	O

Most	O
of	O
the	O
affected	O
devices	O
were	O
located	O
in	O
the	O
Middle	O
East	O
,	O
and	O
many	O
of	O
the	O
stolen	O
data	O
we	O
saw	O
is	O
military-related	O
(	O
e.g.	O
,	O
images	O
,	O
documents	O
)	O
.	O

Bouncing	B-Malware
Golf	I-Malware
’	O
s	O
operators	O
also	O
try	O
to	O
cover	O
their	O
tracks	O
.	O

The	O
registrant	O
contact	O
details	O
of	O
the	O
C	O
&	O
C	O
domains	O
used	O
in	O
the	O
campaign	O
,	O
for	O
instance	O
,	O
were	O
masked	O
.	O

The	O
C	O
&	O
C	O
server	O
IP	O
addresses	O
used	O
also	O
appear	O
to	O
be	O
disparate	O
,	O
as	O
they	O
were	O
located	O
in	O
many	O
European	O
countries	O
like	O
Russia	O
,	O
France	O
,	O
Holland	O
,	O
and	O
Germany	O
.	O

It	O
’	O
s	O
not	O
a	O
definite	O
correlation	O
,	O
but	O
Bouncing	B-Malware
Golf	I-Malware
also	O
seems	O
to	O
have	O
a	O
connection	O
with	O
Domestic	B-Malware
Kitten	I-Malware
due	O
to	O
similarities	O
we	O
found	O
in	O
their	O
code	O
.	O

For	O
example	O
,	O
the	O
Android	B-System
malware	O
that	O
both	O
deploy	O
share	O
the	O
same	O
strings	O
of	O
code	O
for	O
their	O
decoding	O
algorithm	O
.	O

The	O
data	O
that	O
Domestic	B-Malware
Kitten	I-Malware
steals	O
follows	O
a	O
similar	O
format	O
with	O
Bouncing	B-Malware
Golf	I-Malware
’	O
s	O
,	O
with	O
each	O
type	O
of	O
data	O
having	O
a	O
unique	O
identifying	O
character	O
.	O

It	O
’	O
s	O
also	O
worth	O
noting	O
that	O
both	O
campaigns	O
repackage	O
apps	O
that	O
are	O
commonly	O
used	O
in	O
their	O
target	O
’	O
s	O
countries	O
,	O
such	O
as	O
Telegram	B-System
,	O
Kik	B-System
,	O
and	O
Plus	B-System
messaging	O
apps	O
.	O

Figure	O
8	O
.	O

Code	O
snippets	O
showing	O
:	O
the	O
decoding	O
algorithm	O
shared	O
by	O
both	O
Bouncing	B-Malware
Golf	I-Malware
and	O
Domestic	B-Malware
Kitten	I-Malware
(	O
top	O
)	O
,	O
the	O
format	O
of	O
data	O
that	O
Domestic	B-Malware
Kitten	I-Malware
’	O
s	O
malware	O
targets	O
to	O
steal	O
(	O
center	O
)	O
,	O
and	O
how	O
both	O
Bouncing	B-Malware
Golf	I-Malware
(	O
bottom	O
left	O
)	O
and	O
Domestic	B-Malware
Kitten	I-Malware
(	O
bottom	O
right	O
)	O
use	O
"	O
"	O
as	O
a	O
separator	O
in	O
their	O
command	O
strings	O
.	O

As	O
we	O
’	O
ve	O
seen	O
in	O
last	O
year	O
’	O
s	O
mobile	O
threat	O
landscape	O
,	O
we	O
expect	O
more	O
cyberespionage	O
campaigns	O
targeting	O
the	O
mobile	O
platform	O
given	O
its	O
ubiquity	O
,	O
employing	O
tried-and-tested	O
techniques	O
to	O
lure	O
unwitting	O
users	O
.	O

The	O
extent	O
of	O
information	O
that	O
these	O
kinds	O
of	O
threats	O
can	O
steal	O
is	O
also	O
significant	O
,	O
as	O
it	O
lets	O
attackers	O
virtually	O
take	O
over	O
a	O
compromised	O
device	O
.	O

Users	O
should	O
adopt	O
best	O
practices	O
,	O
while	O
organizations	O
should	O
ensure	O
that	O
they	O
balance	O
the	O
need	O
for	O
mobility	O
and	O
the	O
importance	O
of	O
security	O
.	O

End	O
users	O
and	O
enterprises	O
can	O
also	O
benefit	O
from	O
multilayered	O
mobile	O
security	O
solutions	O
such	O
as	O
Trend	B-Organization
Micro™	I-Organization
Mobile	O
Security™	O
.	O

Trend	B-Organization
Micro™	I-Organization
Mobile	B-System
Security	I-System
for	I-System
Enterprise	I-System
provides	O
device	O
,	O
compliance	O
and	O
application	O
management	O
,	O
data	O
protection	O
,	O
and	O
configuration	O
provisioning	O
,	O
as	O
well	O
as	O
protects	O
devices	O
from	O
attacks	O
that	O
exploit	O
vulnerabilities	O
,	O
preventing	O
unauthorized	O
access	O
to	O
apps	O
,	O
and	O
detecting	O
and	O
blocking	O
malware	O
and	O
fraudulent	O
websites	O
.	O

Trend	B-Organization
Micro	I-Organization
’	O
s	O
Mobile	B-System
App	I-System
Reputation	I-System
Service	I-System
(	O
MARS	O
)	O
covers	O
Android	B-System
and	O
iOS	B-System
threats	O
using	O
leading	O
sandbox	O
and	O
machine	O
learning	O
technologies	O
,	O
protecting	O
devices	O
against	O
malware	O
,	O
zero-day	O
and	O
known	O
exploits	O
,	O
privacy	O
leaks	O
,	O
and	O
application	O
vulnerabilities	O
.	O

Several	O
weeks	O
ago	O
,	O
Check	B-Organization
Point	I-Organization
Mobile	O
Threat	O
Prevention	O
detected	O
and	O
quarantined	O
the	O
Android	B-System
device	O
of	O
an	O
unsuspecting	O
customer	O
employee	O
who	O
downloaded	O
and	O
installed	O
a	O
0day	O
mobile	O
ransomware	O
from	O
Google	B-System
Play	I-System
dubbed	O
“	O
Charger.	B-Malware
”	O
This	O
incident	O
demonstrates	O
how	O
malware	O
can	O
be	O
a	O
dangerous	O
threat	O
to	O
your	O
business	O
,	O
and	O
how	O
advanced	O
behavioral	O
detection	O
fills	O
mobile	O
security	O
gaps	O
attackers	O
use	O
to	O
penetrate	O
entire	O
networks	O
.	O

Charger	B-Malware
was	O
found	O
embedded	O
in	O
an	O
app	O
called	O
EnergyRescue	B-Malware
.	O

The	O
infected	O
app	O
steals	O
contacts	O
and	O
SMS	O
messages	O
from	O
the	O
user	O
’	O
s	O
device	O
and	O
asks	O
for	O
admin	O
permissions	O
.	O

If	O
granted	O
,	O
the	O
ransomware	O
locks	O
the	O
device	O
and	O
displays	O
a	O
message	O
demanding	O
payment	O
:	O
You	O
need	O
to	O
pay	O
for	O
us	O
,	O
otherwise	O
we	O
will	O
sell	O
portion	O
of	O
your	O
personal	O
information	O
on	O
black	O
market	O
every	O
30	O
minutes	O
.	O

WE	O
GIVE	O
100	O
%	O
GUARANTEE	O
THAT	O
ALL	O
FILES	O
WILL	O
RESTORE	O
AFTER	O
WE	O
RECEIVE	O
PAYMENT	O
.	O

WE	O
WILL	O
UNLOCK	O
THE	O
MOBILE	O
DEVICE	O
AND	O
DELETE	O
ALL	O
YOUR	O
DATA	O
FROM	O
OUR	O
SERVER	O
!	O

TURNING	O
OFF	O
YOUR	O
PHONE	O
IS	O
MEANINGLESS	O
,	O
ALL	O
YOUR	O
DATA	O
IS	O
ALREADY	O
STORED	O
ON	O
OUR	O
SERVERS	O
!	O

WE	O
STILL	O
CAN	O
SELLING	O
IT	O
FOR	O
SPAM	O
,	O
FAKE	O
,	O
BANK	O
CRIME	O
etc…	O
We	O
collect	O
and	O
download	O
all	O
of	O
your	O
personal	O
data	O
.	O

All	O
information	O
about	O
your	O
social	O
networks	O
,	O
Bank	O
accounts	O
,	O
Credit	O
Cards	O
.	O

We	O
collect	O
all	O
data	O
about	O
your	O
friends	O
and	O
family	O
.	O

The	O
ransom	O
demand	O
for	O
0.2	O
Bitcoins	O
(	O
roughly	O
$	O
180	O
)	O
is	O
a	O
much	O
higher	O
ransom	O
demand	O
than	O
has	O
been	O
seen	O
in	O
mobile	O
ransomware	O
so	O
far	O
.	O

By	O
comparison	O
,	O
the	O
DataLust	B-Malware
ransomware	O
demanded	O
merely	O
$	O
15	O
.	O

Payments	O
are	O
made	O
to	O
a	O
specific	O
Bitcoin	B-System
account	O
,	O
but	O
we	O
haven	O
’	O
t	O
identified	O
any	O
payments	O
so	O
far	O
.	O

Adware	O
commonly	O
found	O
on	O
Play	O
collects	O
profits	O
from	O
ad	O
networks	O
,	O
but	O
mobile	O
ransomware	O
inflicts	O
direct	O
harm	O
to	O
users	O
.	O

Like	O
FakeDefender	B-Malware
and	O
DataLust	B-Malware
,	O
Charger	B-Malware
could	O
be	O
an	O
indicator	O
of	O
a	O
wider	O
effort	O
by	O
mobile	O
malware	O
developers	O
to	O
catch	O
up	O
with	O
their	O
PC	O
ransomware	O
cousins	O
.	O

Similar	O
to	O
other	O
malware	O
seen	O
in	O
the	O
past	O
,	O
Charger	B-Malware
checks	O
the	O
local	O
settings	O
of	O
the	O
device	O
and	O
does	O
not	O
run	O
its	O
malicious	O
logic	O
if	O
the	O
device	O
is	O
located	O
in	O
Ukraine	O
,	O
Russia	O
,	O
or	O
Belarus	O
.	O

This	O
is	O
likely	O
done	O
to	O
keep	O
the	O
developers	O
from	O
being	O
prosecuted	O
in	O
their	O
own	O
countries	O
or	O
being	O
extradited	O
between	O
countries	O
.	O

Most	O
malware	O
found	O
on	O
Google	B-System
Play	I-System
contains	O
only	O
a	O
dropper	O
that	O
later	O
downloads	O
the	O
real	O
malicious	O
components	O
to	O
the	O
device	O
.	O

Charger	B-Malware
,	O
however	O
,	O
uses	O
a	O
heavy	O
packing	O
approach	O
which	O
it	O
harder	O
for	O
the	O
malware	O
to	O
stay	O
hidden	O
,	O
so	O
it	O
must	O
compensate	O
with	O
other	O
means	O
.	O

The	O
developers	O
of	O
Charger	B-Malware
gave	O
it	O
everything	O
they	O
had	O
to	O
boost	O
its	O
evasion	O
capabilities	O
and	O
so	O
it	O
could	O
stay	O
hidden	O
on	O
Google	B-System
Play	I-System
for	O
as	O
long	O
as	O
possible	O
.	O

The	O
malware	O
uses	O
several	O
advanced	O
techniques	O
to	O
hide	O
its	O
real	O
intentions	O
and	O
makes	O
it	O
harder	O
to	O
detect	O
.	O

It	O
encodes	O
strings	O
into	O
binary	O
arrays	O
,	O
making	O
it	O
hard	O
to	O
inspect	O
them	O
.	O

It	O
loads	O
code	O
from	O
encrypted	O
resources	O
dynamically	O
,	O
which	O
most	O
detection	O
engines	O
can	O
not	O
penetrate	O
and	O
inspect	O
.	O

The	O
dynamically-loaded	O
code	O
is	O
also	O
flooded	O
with	O
meaningless	O
commands	O
that	O
mask	O
the	O
actual	O
commands	O
passing	O
through	O
.	O

It	O
checks	O
whether	O
it	O
is	O
being	O
run	O
in	O
an	O
emulator	O
before	O
it	O
starts	O
its	O
malicious	O
activity	O
.	O

PC	O
malware	O
first	O
introduced	O
this	O
technique	O
which	O
is	O
becoming	O
a	O
trend	O
in	O
mobile	O
malware	O
having	O
been	O
adopted	O
by	O
several	O
malware	O
families	O
including	O
Dendroid	B-Malware
.	O

Emulator	O
and	O
location	O
conditions	O
for	O
the	O
malware	O
’	O
s	O
activity	O
Check	B-Organization
Point	I-Organization
Mobile	O
Threat	O
Prevention	O
customers	O
are	O
protected	O
from	O
Charger	B-Malware
and	O
similar	O
malware	O
.	O

Check	B-Organization
Point	I-Organization
’	O
s	O
Analysis	O
and	O
Response	O
Team	O
(	O
ART	O
)	O
disclosed	O
the	O
finding	O
to	O
Android	B-System
’	O
s	O
Security	O
team	O
who	O
took	O
the	O
appropriate	O
security	O
steps	O
to	O
remove	O
the	O
infected	O
app	O
and	O
added	O
the	O
malware	O
to	O
Android	B-System
’	O
s	O
built-in	O
protection	O
mechanisms	O
.	O

Charger	O
SHA256	O
hash	O
:	O
58eb6c368e129b17559bdeacb3aed4d9a5d3596f774cf5ed3fdcf51775232ba0	B-Indicator
Infostealer	O
,	O
Keylogger	O
,	O
and	O
Ransomware	O
in	O
One	O
:	O
Anubis	B-Malware
Targets	O
More	O
than	O
250	O
Android	B-System
Applications	O
October	O
29	O
,	O
2021	O
The	O
Cofense	B-Organization
Phishing	I-Organization
Defense	I-Organization
Center	I-Organization
uncovered	O
a	O
phishing	O
campaign	O
that	O
specifically	O
targets	O
users	O
of	O
Android	B-System
devices	O
that	O
could	O
result	O
in	O
compromise	O
if	O
unsigned	O
Android	B-System
applications	O
are	O
permitted	O
on	O
the	O
device	O
.	O

The	O
campaign	O
seeks	O
to	O
deliver	O
Anubis	B-Malware
,	O
a	O
particularly	O
nasty	O
piece	O
of	O
malware	O
that	O
was	O
originally	O
used	O
for	O
cyber	O
espionage	O
and	O
retooled	O
as	O
a	O
banking	O
trojan	O
.	O

Anubis	B-Malware
can	O
completely	O
hijack	O
an	O
Android	B-System
mobile	O
device	O
,	O
steal	O
data	O
,	O
record	O
phone	O
calls	O
,	O
and	O
even	O
hold	O
the	O
device	O
to	O
ransom	O
by	O
encrypting	O
the	O
victim	O
’	O
s	O
personal	O
files	O
.	O

With	O
mobile	O
devices	O
increasingly	O
used	O
in	O
the	O
corporate	O
environment	O
,	O
thanks	O
to	O
the	O
popularity	O
of	O
BYOD	O
policies	O
,	O
this	O
malware	O
has	O
the	O
potential	O
to	O
cause	O
serious	O
harm	O
,	O
mostly	O
to	O
consumers	O
,	O
and	O
businesses	O
that	O
allow	O
the	O
installation	O
of	O
unsigned	O
applications	O
.	O

Here	O
’	O
s	O
how	O
it	O
works	O
:	O
At	O
first	O
glance	O
,	O
the	O
email	O
shown	O
in	O
Figure	O
1	O
looks	O
like	O
any	O
other	O
phishing	O
email	O
that	O
asks	O
the	O
user	O
to	O
download	O
an	O
invoice	O
.	O

However	O
,	O
this	O
particular	O
email	O
downloads	O
an	O
Android	B-System
Package	I-System
Kit	I-System
(	O
APK	O
)	O
,	O
which	O
is	O
the	O
common	O
format	O
used	O
by	O
Android	B-System
to	O
distribute	O
and	O
install	O
applications	O
.	O

Let	O
’	O
s	O
take	O
a	O
closer	O
look	O
at	O
the	O
suspicious	O
file	O
.	O

Figure	O
1	O
–	O
Phishing	O
Email	O
When	O
the	O
email	O
link	O
is	O
opened	O
from	O
an	O
Android	B-System
device	O
,	O
an	O
APK	O
file	O
(	O
Fattura002873.apk	B-Indicator
)	O
,	O
is	O
downloaded	O
.	O

Upon	O
opening	O
the	O
file	O
,	O
the	O
user	O
is	O
asked	O
to	O
enable	O
“	O
Google	B-System
Play	I-System
Protect	O
”	O
as	O
shown	O
in	O
Figure	O
2	O
.	O

However	O
,	O
this	O
is	O
not	O
a	O
genuine	O
“	O
Google	B-System
Play	I-System
Protect	O
”	O
screen	O
;	O
instead	O
it	O
gives	O
the	O
app	O
all	O
the	O
permissions	O
it	O
needs	O
while	O
simultaneously	O
disabling	O
the	O
actual	O
Google	B-System
Play	I-System
Protect	I-System
.	O

Figure	O
2	O
–	O
Granting	O
Permissions	O
The	O
following	O
permissions	O
are	O
granted	O
to	O
the	O
app	O
:	O
Figure	O
3	O
–	O
Permissions	O
Granted	O
to	O
App	O
A	O
closer	O
look	O
at	O
the	O
code	O
reveals	O
the	O
application	O
gathers	O
a	O
list	O
of	O
installed	O
applications	O
to	O
compare	O
the	O
results	O
against	O
a	O
list	O
of	O
targeted	O
applications	O
(	O
Figure	O
4	O
)	O
.	O

The	O
malware	O
mainly	O
targets	O
banking	O
and	O
financial	O
applications	O
,	O
but	O
also	O
looks	O
for	O
popular	O
shopping	O
apps	O
such	O
as	O
eBay	B-Organization
or	O
Amazon	B-Organization
.	O

A	O
full	O
list	O
of	O
targeted	O
applications	O
is	O
included	O
in	O
the	O
IOC	O
section	O
at	O
the	O
end	O
of	O
this	O
post	O
.	O

Once	O
an	O
application	O
has	O
been	O
identified	O
,	O
Anubis	B-Malware
overlays	O
the	O
original	O
application	O
with	O
a	O
fake	O
login	O
page	O
to	O
capture	O
the	O
user	O
’	O
s	O
credentials	O
.	O

Figure	O
4	O
–	O
Checking	O
for	O
installed	O
apps	O
Based	O
on	O
a	O
thorough	O
analysis	O
of	O
the	O
code	O
,	O
the	O
most	O
interesting	O
technical	O
capabilities	O
include	O
:	O
Capturing	O
screenshots	O
Enabling	O
or	O
changing	O
administration	O
settings	O
Opening	O
and	O
visiting	O
any	O
URL	O
Disabling	O
Play	O
Protect	O
Recording	O
audio	O
Making	O
phone	O
calls	O
Stealing	O
the	O
contact	O
list	O
Controlling	O
the	O
device	O
via	O
VNC	O
Sending	O
,	O
receiving	O
and	O
deleting	O
SMS	O
Locking	O
the	O
device	O
Encrypting	O
files	O
on	O
the	O
device	O
and	O
external	O
drives	O
Searching	O
for	O
files	O
Retrieving	O
the	O
GPS	O
location	O
Capturing	O
remote	O
control	O
commands	O
from	O
Twitter	B-System
and	O
Telegram	B-System
Pushing	O
overlays	O
Reading	O
the	O
device	O
ID	O
The	O
malware	O
includes	O

a	O
keylogger	O
that	O
works	O
in	O
every	O
app	O
installed	O
on	O
the	O
Android	B-System
device	O
.	O

However	O
,	O
the	O
keylogger	O
needs	O
to	O
be	O
specifically	O
enabled	O
by	O
a	O
command	O
sent	O
from	O
the	O
C2	O
server	O
.	O

The	O
keylogger	O
can	O
track	O
three	O
different	O
events	O
(	O
Figure	O
5	O
)	O
:	O
TYPE_VIEW_CLICKED	O
Represents	O
the	O
event	O
of	O
clicking	O
on	O
a	O
View-like	O
Button	O
,	O
CompoundButton	O
,	O
etc	O
.	O

TYPE_VIEW_FOCUSED	O
Represents	O
the	O
event	O
of	O
setting	O
input	O
focus	O
of	O
a	O
View	O
.	O

TYPE_VIEW_TEXT_CHANGED	O
Represents	O
the	O
event	O
of	O
changing	O
the	O
text	O
of	O
an	O
EditText	O
.	O

Figure	O
5	O
–	O
Keylogger	O
component	O
Figure	O
6	O
shows	O
one	O
of	O
the	O
most	O
noteworthy	O
functions	O
of	O
Anubis	B-Malware
:	O
its	O
ransomware	O
module	O
.	O

The	O
malware	O
searches	O
both	O
internal	O
and	O
external	O
storage	O
and	O
encrypts	O
them	O
using	O
RC4	O
.	O

It	O
adds	O
the	O
file	O
extension	O
.AnubisCrypt	B-Indicator
to	O
each	O
encrypted	O
file	O
and	O
sends	O
it	O
to	O
the	O
C2	O
.	O

Figure	O
6	O
–	O
Ransomware	O
component	O
Anubis	B-Malware
has	O
been	O
known	O
to	O
utilize	O
Twitter	B-Organization
or	O
Telegram	B-Organization
to	O
retrieve	O
the	O
C2	O
address	O
and	O
this	O
sample	O
is	O
no	O
exception	O
(	O
Figure	O
7	O
)	O
.	O

Figure	O
7	O
–	O
C2	O
As	O
seen	O
in	O
Figure	O
8	O
,	O
this	O
version	O
of	O
Anubis	B-Malware
is	O
built	O
to	O
run	O
on	O
several	O
iterations	O
of	O
the	O
Android	B-System
operating	O
system	O
,	O
dating	O
back	O
to	O
version	O
4.0.3	O
,	O
which	O
was	O
released	O
in	O
2012	O
.	O

Figure	O
8	O
–	O
Android	B-System
requirements	O
Android	B-System
malware	O
has	O
been	O
around	O
for	O
many	O
years	O
and	O
will	O
be	O
with	O
us	O
for	O
the	O
foreseeable	O
future	O
.	O

Users	O
who	O
have	O
configured	O
their	O
Android	B-System
mobile	O
device	O
to	O
receive	O
work-related	O
emails	O
and	O
allow	O
installation	O
of	O
unsigned	O
applications	O
face	O
the	O
most	O
risk	O
of	O
compromise	O
.	O

APK	O
files	O
will	O
not	O
natively	O
open	O
in	O
an	O
environment	O
other	O
than	O
an	O
Android	B-System
device	O
.	O

With	O
the	O
increased	O
use	O
of	O
Android	B-System
phones	O
in	O
business	O
environments	O
,	O
it	O
is	O
important	O
to	O
defend	O
against	O
these	O
threats	O
by	O
ensuring	O
devices	O
are	O
kept	O
current	O
with	O
the	O
latest	O
updates	O
.	O

Limiting	O
app	O
installations	O
on	O
corporate	O
devices	O
,	O
as	O
well	O
as	O
ensuring	O
that	O
applications	O
are	O
created	O
by	O
trusted	O
developers	O
on	O
official	O
marketplaces	O
,	O
can	O
help	O
in	O
reducing	O
the	O
risk	O
of	O
infection	O
as	O
well	O
.	O

ViceLeaker	B-Malware
Operation	O
:	O
mobile	O
espionage	O
targeting	O
Middle	O
East	O
26	O
JUN	O
2019	O
In	O
May	O
2018	O
,	O
we	O
discovered	O
a	O
campaign	O
targeting	O
dozens	O
of	O
mobile	O
Android	B-System
devices	O
belonging	O
to	O
Israeli	O
citizens	O
.	O

Kaspersky	B-Organization
spyware	O
sensors	O
caught	O
the	O
signal	O
of	O
an	O
attack	O
from	O
the	O
device	O
of	O
one	O
of	O
the	O
victims	O
;	O
and	O
a	O
hash	O
of	O
the	O
APK	O
involved	O
(	O
Android	B-System
application	O
)	O
was	O
tagged	O
in	O
our	O
sample	O
feed	O
for	O
inspection	O
.	O

Once	O
we	O
looked	O
into	O
the	O
file	O
,	O
we	O
quickly	O
found	O
out	O
that	O
the	O
inner-workings	O
of	O
the	O
APK	O
included	O
a	O
malicious	O
payload	O
,	O
embedded	O
in	O
the	O
original	O
code	O
of	O
the	O
application	O
.	O

This	O
was	O
an	O
original	O
spyware	O
program	O
,	O
designed	O
to	O
exfiltrate	O
almost	O
all	O
accessible	O
information	O
.	O

During	O
the	O
course	O
of	O
our	O
research	O
,	O
we	O
noticed	O
that	O
we	O
were	O
not	O
the	O
only	O
ones	O
to	O
have	O
found	O
the	O
operation	O
.	O

Researchers	O
from	O
Bitdefender	B-System
also	O
released	O
an	O
analysis	O
of	O
one	O
of	O
the	O
samples	O
in	O
a	O
blogpost	O
.	O

Although	O
something	O
had	O
already	O
been	O
published	O
,	O
we	O
decided	O
to	O
do	O
something	O
different	O
with	O
the	O
data	O
we	O
acquired	O
.	O

The	O
following	O
month	O
,	O
we	O
released	O
a	O
private	O
report	O
on	O
our	O
Threat	O
Intelligence	O
Portal	O
to	O
alert	O
our	O
clients	O
about	O
this	O
newly	O
discovered	O
operation	O
and	O
began	O
writing	O
YARA	O
rules	O
in	O
order	O
to	O
catch	O
more	O
samples	O
.	O

We	O
decided	O
to	O
call	O
the	O
operation	O
“	O
ViceLeaker	B-Malware
”	O
,	O
because	O
of	O
strings	O
and	O
variables	O
in	O
its	O
code	O
.	O

Mobile	O
ViceLeaker	B-Malware
The	O
following	O
table	O
shows	O
meta	O
information	O
on	O
the	O
observed	O
samples	O
,	O
including	O
compiler	O
timestamps	O
:	O
MD5	O
Package	O
Compiler	O
C2	O
51df2597faa3fce38a4c5ae024f97b1c	B-Indicator
com.xapps.SexGameForAdults	B-Indicator
dexlib	O
2.x	O
188.165.28	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
251	I-Indicator
2d108ff3a735dea1d1fdfa430f37fab2	B-Indicator
com.psiphon3	B-Indicator
dexlib	O
2.x	O
188.165.49	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
205	I-Indicator
7ed754a802f0b6a1740a99683173db73	B-Indicator
com.psiphon3	B-Indicator
dexlib	O
2.x	O
188.165.49	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
205	I-Indicator
3b89e5cd49c05ce6dc681589e6c368d9	B-Indicator
ir.abed.dastan	B-Indicator
dexlib	O
2.x	O
185.141.60	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
213	I-Indicator
To	O
backdoor	O
legitimate	O
applications	O
,	O
attackers	O
used	O
a	O
Smali	O
injection	O
technique	O
–	O
a	O
type	O
of	O
injection	O
that	O
allows	O
attackers	O
to	O
disassemble	O
the	O
code	O
of	O
original	O
app	O
with	O
the	O
Baksmali	O
tool	O
,	O
add	O
their	O
malicious	O
code	O
,	O
and	O
assemble	O
it	O
with	O
Smali	O
.	O

As	O
a	O
result	O
,	O
due	O
to	O
such	O
an	O
unusual	O
compilation	O
process	O
,	O
there	O
were	O
signs	O
in	O
the	O
dex	O
file	O
that	O
point	O
to	O
dexlib	O
,	O
a	O
library	O
used	O
by	O
the	O
Smali	O
tool	O
to	O
assemble	O
dex	O
files	O
.	O

Original	O
code	O
of	O
the	O
APK	O
on	O
the	O
left	O
,	O
versus	O
injected	O
APK	O
on	O
the	O
right	O
The	O
analysis	O
of	O
the	O
APK	O
was	O
rather	O
interesting	O
,	O
because	O
some	O
of	O
the	O
actions	O
were	O
very	O
common	O
spyware	O
features	O
,	O
such	O
as	O
the	O
exfiltration	O
of	O
SMS	O
messages	O
,	O
call	O
logs	O
and	O
other	O
data	O
.	O

However	O
,	O
in	O
addition	O
to	O
the	O
traditional	O
functionality	O
,	O
there	O
were	O
also	O
backdoor	O
capabilities	O
such	O
as	O
upload	O
,	O
download	O
,	O
delete	O
files	O
,	O
camera	O
takeover	O
and	O
record	O
surrounding	O
audio	O
.	O

The	O
malware	O
uses	O
HTTP	O
for	O
communication	O
with	O
the	O
C2	O
server	O
for	O
command	O
handling	O
and	O
data	O
exfiltration	O
.	O

Here	O
is	O
a	O
command	O
and	O
control	O
protocol	O
fragment	O
:	O
Commands	O
from	O
C2	O
server	O
parsing	O
In	O
total	O
,	O
the	O
malicious	O
APK	O
handles	O
16	O
different	O
commands	O
:	O
Command	O
Endpoint	O
Description	O
1	O
reqsmscal.php	B-Indicator
Send	O
specified	O
SMS	O
message	O
2	O
reqsmscal.php	B-Indicator
Call	O
specified	O
number	O
3	O
reqsmscal.php	B-Indicator
Exfiltrate	O
device	O
info	O
,	O
such	O
as	O
phone	O
model	O
and	O
OS	O
version	O
4	O
reqsmscal.php	B-Indicator
Exfiltrate	O
a	O
list	O
of	O
all	O
installed	O
applications	O
5	O
reqsmscal.php	B-Indicator
Exfiltrate	O
default	O
browser	O
history	O
(	O
limited	O
to	O
a	O
given	O
date	O
)	O
6	O
reqsmscal.php	B-Indicator

Exfiltrate	O
Chrome	O
browser	O
history	O
(	O
limited	O
to	O
a	O
given	O
date	O
)	O
7	O
reqsmscal.php	B-Indicator
Exfiltrate	O
memory	O
card	O
file	O
structure	O
8	O
reqsmscal.php	B-Indicator
Record	O
surrounding	O
sound	O
for	O
80	O
seconds	O
1	O
reqcalllog.php	B-Indicator
Exfiltrate	O
all	O
call	O
logs	O
2	O
reqcalllog.php	B-Indicator
Exfiltrate	O
all	O
SMS	O
messages	O
3	O
reqcalllog.php	B-Indicator
Upload	O
specified	O
file	O
from	O
the	O
device	O
to	O
the	O
C2	O
4	O
reqcalllog.php	B-Indicator
Download	O
file	O
from	O
specified	O
URL	O
and	O
save	O
on	O
device	O
5	O
reqcalllog.php	B-Indicator
Delete	O
specified	O
file	O
6,7,8	O
reqcalllog.php	B-Indicator
Commands	O
not	O
yet	O

implemented	O
9	O
reqcalllog.php	B-Indicator
Take	O
photo	O
(	O
muted	O
audio	O
)	O
with	O
rear	O
camera	O
,	O
send	O
to	O
C2	O
10	O
reqcalllog.php	B-Indicator
Take	O
photo	O
(	O
muted	O
audio	O
)	O
with	O
front	O
camera	O
,	O
send	O
to	O
C2	O
All	O
observed	O
samples	O
with	O
Smali	O
injections	O
were	O
signed	O
by	O
the	O
same	O
debug	O
certificate	O
(	O
0x936eacbe07f201df	O
)	O
.	O

As	O
we	O
know	O
from	O
our	O
investigation	O
,	O
traces	O
of	O
the	O
first	O
development	O
activities	O
were	O
found	O
at	O
the	O
end	O
of	O
2016	O
,	O
but	O
the	O
main	O
distribution	O
campaign	O
began	O
in	O
2018	O
(	O
end	O
of	O
2017	O
)	O
.	O

Based	O
on	O
our	O
detection	O
statistics	O
,	O
the	O
main	O
infection	O
vector	O
is	O
the	O
spread	O
of	O
Trojanized	O
applications	O
directly	O
to	O
victims	O
via	O
Telegram	O
and	O
WhatsApp	O
messengers	O
.	O

There	O
are	O
the	O
following	O
relevant	O
detection	O
paths	O
(	O
the	O
last	O
one	O
is	O
an	O
alternative	O
Telegram	O
client	O
–	O
“	O
Telegram	O
X	O
“	O
)	O
:	O
Name	O
Detection	O
path	O
Sex	O
Game	O
For	O
Adults	O
18.apk	B-Indicator
/storage/emulated/0/WhatsApp/Media/WhatsApp	B-Indicator
Documents/	I-Indicator
4_6032967490689041387.apk	I-Indicator
/storage/emulated/0/Telegram/Telegram	I-Indicator
Documents/	I-Indicator
Psiphon-v91.apk	B-Indicator
/storage/emulated/0/Android/data/org.thunderdog.challegram/files/documents/	B-Indicator
Backdoored	O
Open	O
Source	O
During	O
the	O
course	O

of	O
our	O
analysis	O
,	O
we	O
also	O
found	O
samples	O
sharing	O
code	O
with	O
the	O
ViceLeaker	B-Malware
malware	O
,	O
in	O
particular	O
they	O
shared	O
a	O
delimiter	O
that	O
was	O
used	O
in	O
both	O
cases	O
to	O
parse	O
commands	O
from	O
the	O
C2	O
server	O
.	O

This	O
would	O
be	O
a	O
very	O
unusual	O
coincidence	O
.	O

Even	O
when	O
a	O
false	O
flag	O
might	O
also	O
be	O
a	O
possibility	O
,	O
we	O
consider	O
this	O
to	O
be	O
unlikely	O
.	O

The	O
samples	O
sharing	O
this	O
overlap	O
are	O
modified	O
versions	O
of	O
an	O
open	O
source	O
Jabber/XMPP	B-System
client	O
called	O
“	O
Conversations	O
”	O
with	O
some	O
code	O
additions	O
.	O

The	O
legitimate	O
version	O
of	O
this	O
app	O
is	O
also	O
available	O
on	O
Google	B-System
Play	I-System
.	O

The	O
Conversations	O
modified	O
samples	O
differ	O
from	O
the	O
original	O
one	O
in	O
the	O
getKnownHosts	O
method	O
that	O
was	O
modified	O
to	O
replace	O
the	O
main	O
XMPP	B-System
host	O
with	O
the	O
attackers	O
’	O
C2	O
server	O
:	O
It	O
appears	O
that	O
the	O
attackers	O
were	O
using	O
a	O
specific	O
C2	O
for	O
the	O
use	O
of	O
that	O
app	O
.	O

Another	O
important	O
modification	O
is	O
in	O
the	O
message	O
transfer	O
process	O
:	O
With	O
this	O
modification	O
,	O
an	O
application	O
sends	O
device	O
location	O
coordinates	O
with	O
every	O
message	O
.	O

There	O
are	O
also	O
many	O
other	O
modifications	O
,	O
fully	O
described	O
in	O
our	O
private	O
report	O
.	O

In	O
addition	O
,	O
we	O
did	O
not	O
see	O
traces	O
of	O
the	O
Smali	O
injection	O
.	O

In	O
this	O
case	O
we	O
found	O
traces	O
of	O
dx/dexmerge	O
compilers	O
,	O
which	O
means	O
that	O
,	O
this	O
time	O
,	O
the	O
attackers	O
just	O
imported	O
the	O
original	O
source	O
code	O
into	O
an	O
Android	B-System
IDE	O
(	O
such	O
as	O
Android	B-System
Studio	I-System
,	O
for	O
instance	O
)	O
and	O
compiled	O
it	O
with	O
their	O
own	O
modifications	O
.	O

In	O
addition	O
to	O
adding	O
the	O
code	O
,	O
the	O
attackers	O
also	O
changed	O
the	O
icon	O
and	O
package	O
name	O
.	O

We	O
do	O
not	O
know	O
why	O
,	O
but	O
we	O
suspect	O
that	O
it	O
was	O
an	O
attempt	O
to	O
hide	O
the	O
origin	O
of	O
the	O
application	O
.	O

Conversations-based	O
app	O
mimics	O
Telegram	B-System
messenger	I-System
Even	O
when	O
we	O
originally	O
thought	O
this	O
was	O
a	O
backdoored	O
version	O
of	O
the	O
Conversations	O
app	O
,	O
used	O
to	O
infect	O
victims	O
,	O
we	O
didn´t	O
discovered	O
anything	O
malicious	O
in	O
it	O
.	O

This	O
brought	O
to	O
us	O
the	O
hypothesis	O
that	O
this	O
might	O
be	O
a	O
version	O
used	O
by	O
the	O
group	O
behind	O
ViceLeaker	B-Malware
for	O
internal	O
communication	O
or	O
for	O
other	O
,	O
unclear	O
purposes	O
.	O

All	O
the	O
detections	O
of	O
this	O
backdoored	O
app	O
were	O
geolocated	O
in	O
Iran	O
.	O

Backdoored	O
Conversations	O
C2	O
server	O
analysis	O
During	O
the	O
analysis	O
of	O
the	O
Smali	O
injected	O
apps	O
and	O
their	O
C2	O
server	O
infrastructure	O
we	O
hadn	O
’	O
t	O
found	O
any	O
interesting	O
clues	O
,	O
but	O
things	O
changed	O
when	O
we	O
looked	O
at	O
the	O
C2	O
server	O
of	O
the	O
linked	O
Conversations	O
messenger	O
.	O

It	O
uses	O
“	O
185.51.201	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
133	I-Indicator
”	O
as	O
a	O
main	O
C2	O
address	O
,	O
and	O
there	O
is	O
only	O
one	O
domain	O
that	O
is	O
hosted	O
on	O
this	O
dedicated	O
server	O
–	O
iliageram	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
ir	I-Indicator
.	O

Note	O
that	O
we	O
later	O
found	O
versions	O
that	O
used	O
the	O
domain	O
as	O
a	O
C2	O
directly	O
instead	O
of	O
the	O
IP	O
address	O
.	O

The	O
record	O
contains	O
a	O
personal	O
email	O
address	O
:	O
WHOIS	O
records	O
of	O
C2	O
server	O
exposing	O
the	O
attacker	O
’	O
s	O
email	O
address	O
We	O
were	O
aware	O
of	O
the	O
possibility	O
that	O
the	O
attackers	O
might	O
be	O
using	O
a	O
compromised	O
email	O
account	O
,	O
so	O
we	O
dug	O
deeper	O
to	O
find	O
more	O
information	O
related	O
to	O
this	O
email	O
address	O
.	O

A	O
quick	O
search	O
produced	O
results	O
about	O
a	O
personal	O
page	O
and	O
,	O
what	O
is	O
more	O
interesting	O
,	O
a	O
GitHub	B-Organization
account	O
that	O
contains	O
a	O
forked	O
Conversation	O
repository	O
.	O

Related	O
Github	B-Organization
account	O
contains	O
forked	O
Conversations	O
repository	O
Summarizing	O
all	O
the	O
found	O
clues	O
,	O
we	O
have	O
the	O
following	O
attribution	O
flow	O
:	O
Conclusion	O
The	O
operation	O
of	O
ViceLeaker	B-Malware
is	O
still	O
ongoing	O
,	O
as	O
is	O
our	O
research	O
.	O

The	O
attackers	O
have	O
taken	O
down	O
their	O
communication	O
channels	O
and	O
are	O
probably	O
looking	O
for	O
ways	O
to	O
assemble	O
their	O
tools	O
in	O
a	O
different	O
manner	O
.	O

Kaspersky	B-Organization
detects	O
and	O
blocks	O
samples	O
of	O
the	O
ViceLeaker	B-Malware
operation	O
using	O
the	O
following	O
verdict	O
:	O
Trojan-Spy.AndroidOS.ViceLeaker	B-Indicator
.	I-Indicator

*	I-Indicator
Actually	O
,	O
we	O
are	O
currently	O
investigating	O
whether	O
this	O
group	O
might	O
also	O
be	O
behind	O
a	O
large-scale	O
web-oriented	O
attack	O
at	O
the	O
end	O
of	O
2018	O
using	O
code	O
injection	O
and	O
exploiting	O
SQL	B-Vulnerability
vulnerabilities	I-Vulnerability
.	O

Even	O
when	O
this	O
would	O
not	O
be	O
directly	O
related	O
to	O
the	O
Android	B-System
malware	O
described	O
in	O
this	O
blogpost	O
,	O
it	O
would	O
be	O
an	O
indicator	O
of	O
wider	O
capabilities	O
and	O
objectives	O
of	O
this	O
actor	O
.	O

XLoader	B-Malware
Android	B-System
Spyware	O
and	O
Banking	O
Trojan	O
Distributed	O
via	O
DNS	O
Spoofing	O
We	O
have	O
been	O
detecting	O
a	O
new	O
wave	O
of	O
network	O
attacks	O
since	O
early	O
March	O
,	O
which	O
,	O
for	O
now	O
,	O
are	O
targeting	O
Japan	O
,	O
Korea	O
,	O
China	O
,	O
Taiwan	O
,	O
and	O
Hong	O
Kong	O
.	O

Trend	B-Organization
Micro	I-Organization
detects	O
these	O
as	O
ANDROIDOS_XLOADER.HRX	B-Indicator
.	O

By	O
:	O
Trend	B-Organization
Micro	I-Organization
April	O
20	O
,	O
2018	O
We	O
have	O
been	O
detecting	O
a	O
new	O
wave	O
of	O
network	O
attacks	O
since	O
early	O
March	O
,	O
which	O
,	O
for	O
now	O
,	O
are	O
targeting	O
Japan	O
,	O
Korea	O
,	O
China	O
,	O
Taiwan	O
,	O
and	O
Hong	O
Kong	O
.	O

The	O
attacks	O
use	O
Domain	O
Name	O
System	O
(	O
DNS	O
)	O
cache	O
poisoning/DNS	O
spoofing	O
,	O
possibly	O
through	O
infringement	O
techniques	O
such	O
as	O
brute-force	O
or	O
dictionary	O
attacks	O
,	O
to	O
distribute	O
and	O
install	O
malicious	O
Android	B-System
apps	O
.	O

Trend	B-Organization
Micro	I-Organization
detects	O
these	O
as	O
ANDROIDOS_XLOADER.HRX	B-Indicator
.	O

These	O
malware	O
pose	O
as	O
legitimate	O
Facebook	B-System
or	O
Chrome	B-System
applications	O
.	O

They	O
are	O
distributed	O
from	O
polluted	O
DNS	O
domains	O
that	O
send	O
a	O
notification	O
to	O
an	O
unknowing	O
victim	O
’	O
s	O
device	O
.	O

The	O
malicious	O
apps	O
can	O
steal	O
personally	O
identifiable	O
and	O
financial	O
data	O
and	O
install	O
additional	O
apps	O
.	O

XLoader	B-Malware
can	O
also	O
hijack	O
the	O
infected	O
device	O
(	O
i.e.	O
,	O
send	O
SMSs	O
)	O
and	O
sports	O
self-protection/persistence	O
mechanisms	O
through	O
device	O
administrator	O
privileges	O
.	O

Infection	O
Chain	O
As	O
with	O
our	O
earlier	O
reports	O
in	O
late	O
March	O
,	O
the	O
attack	O
chain	O
involves	O
diverting	O
internet	O
traffic	O
to	O
attacker-specified	O
domains	O
by	O
compromising	O
and	O
overwriting	O
the	O
router	O
’	O
s	O
DNS	O
settings	O
.	O

A	O
fake	O
alert	O
will	O
notify	O
and	O
urge	O
the	O
user	O
to	O
access	O
the	O
malicious	O
domain	O
and	O
download	O
XLoader	B-Malware
.	O

Technical	O
Analysis	O
XLoader	B-Malware
first	O
loads	O
the	O
encrypted	O
payload	O
from	O
Assets/db	B-Indicator
as	O
test.dex	B-Indicator
to	O
drop	O
the	O
necessary	O
modules	O
then	O
requests	O
for	O
device	O
administrator	O
privileges	O
.	O

Once	O
granted	O
permission	O
,	O
it	O
hides	O
its	O
icon	O
from	O
the	O
launcher	O
application	O
list	O
then	O
starts	O
a	O
service	O
that	O
it	O
keeps	O
running	O
in	O
the	O
background	O
.	O

The	O
background	O
service	O
uses	O
the	O
reflection	O
technique	O
(	O
a	O
feature	O
that	O
allows	O
the	O
inspection	O
and	O
modification	O
of	O
Java-based	O
programs	O
’	O
internal	O
properties	O
)	O
to	O
invoke	O
the	O
method	O
com.Loader.start	B-Indicator
in	O
the	O
payload	O
.	O

Monitoring	O
Broadcast	O
Events	O
XLoader	B-Malware
registers	O
many	O
broadcast	O
receivers	O
in	O
the	O
payload	O
dynamically	O
(	O
to	O
monitor	O
broadcast	O
events	O
sent	O
between	O
system	O
and	O
applications	O
)	O
.	O

Registering	O
broadcast	O
receivers	O
enable	O
XLoader	B-Malware
to	O
trigger	O
its	O
malicious	O
routines	O
.	O

Here	O
is	O
a	O
list	O
of	O
broadcast	O
actions	O
:	O
android.provider.Telephony.SMS_RECEIVED	B-Indicator
android.net.conn.CONNECTIVITY_CHANGE	B-Indicator
android.intent.action.BATTERY_CHANGED	B-Indicator
android.intent.action.USER_PRESENT	B-Indicator
android.intent.action.PHONE_STATE	B-Indicator
android.net.wifi.SCAN_RESULTS	B-Indicator
android.intent.action.PACKAGE_ADDED	B-Indicator
android.intent.action.PACKAGE_REMOVED	B-Indicator
android.intent.action.SCREEN_OFF	B-Indicator
android.intent.action.SCREEN_ON	B-Indicator

android.media.RINGER_MODE_CHANGED	B-Indicator
android.sms.msg.action.SMS_SEND	B-Indicator
android.sms.msg.action.SMS_DELIVERED	B-Indicator
Creating	O
a	O
Web	O
Server	O
to	O
Phish	O
XLoader	B-Malware
creates	O
a	O
provisional	O
web	O
server	O
to	O
receive	O
the	O
broadcast	O
events	O
.	O

It	O
can	O
also	O
create	O
a	O
simple	O
HTTP	O
server	O
on	O
the	O
infected	O
device	O
to	O
deceive	O
victims	O
.	O

It	O
shows	O
a	O
web	O
phishing	O
page	O
whenever	O
the	O
affected	O
device	O
receives	O
a	O
broadcast	O
event	O
(	O
i.e.	O
,	O
if	O
a	O
new	O
package	O
is	O
installed	O
or	O
if	O
the	O
device	O
’	O
s	O
screen	O
is	O
on	O
)	O
to	O
steal	O
personal	O
data	O
,	O
such	O
as	O
those	O
keyed	O
in	O
for	O
banking	O
apps	O
.	O

The	O
phishing	O
page	O
is	O
translated	O
in	O
Korean	O
,	O
Japanese	O
,	O
Chinese	O
,	O
and	O
English	O
,	O
which	O
are	O
hardcoded	O
in	O
the	O
payload	O
.	O

It	O
will	O
appear	O
differently	O
to	O
users	O
depending	O
on	O
the	O
language	O
set	O
on	O
the	O
device	O
.	O

XLoader	B-Malware
as	O
Spyware	O
and	O
Banking	O
Trojan	O
XLoader	B-Indicator
can	O
also	O
collect	O
information	O
related	O
to	O
usage	O
of	O
apps	O
installed	O
in	O
the	O
device	O
.	O

Its	O
data-stealing	O
capabilities	O
include	O
collecting	O
SMSs	O
after	O
receiving	O
an	O
SMS-related	O
broadcast	O
event	O
and	O
covertly	O
recording	O
phone	O
calls	O
.	O

XLoader	B-Malware
can	O
also	O
hijack	O
accounts	O
linked	O
to	O
financial	O
or	O
game-related	O
apps	O
installed	O
on	O
the	O
affected	O
device	O
.	O

XLoader	B-Malware
can	O
also	O
start	O
other	O
attacker-specified	O
packages	O
.	O

A	O
possible	O
attack	O
scenario	O
involves	O
replacing	O
legitimate	O
apps	O
with	O
repackaged	O
or	O
malicious	O
versions	O
.	O

By	O
monitoring	O
the	O
package	O
installation	O
broadcast	O
event	O
,	O
XLoader	B-Malware
can	O
start	O
their	O
packages	O
.	O

This	O
enables	O
it	O
to	O
launch	O
malicious	O
apps	O
without	O
the	O
user	O
’	O
s	O
awareness	O
and	O
explicit	O
consent	O
.	O

We	O
reverse	O
engineered	O
XLoader	B-Malware
and	O
found	O
that	O
it	O
appears	O
to	O
target	O
South	O
Korea-based	O
banks	O
and	O
game	O
development	O
companies	O
.	O

XLoader	B-Malware
also	O
prevents	O
victims	O
from	O
accessing	O
the	O
device	O
’	O
s	O
settings	O
or	O
using	O
a	O
known	O
antivirus	O
(	O
AV	O
)	O
app	O
in	O
the	O
country	O
.	O

XLoader	B-Malware
can	O
also	O
load	O
multiple	O
malicious	O
modules	O
to	O
receive	O
and	O
execute	O
commands	O
from	O
its	O
remote	O
command-and-control	O
(	O
C	O
&	O
C	O
)	O
server	O
,	O
as	O
shown	O
below	O
:	O
Here	O
’	O
s	O
a	O
list	O
of	O
the	O
modules	O
and	O
their	O
functions	O
:	O
sendSms	O
—	O
send	O
SMS/MMS	O
to	O
a	O
specified	O
address	O
setWifi	O
—	O
enable	O
or	O
disable	O
Wi-Fi	O
connection	O
gcont	O
—	O
collect	O
all	O
the	O
device	O
’	O
s	O
contacts	O
lock	O
—	O
currently	O
just	O
an	O
input	O
lock	O
status	O
in	O
the	O
settings	O
(	O
pref	O
)	O
file	O
,	O
but	O
may	O
be	O
used	O
as	O
a	O
screenlocking	O
ransomware	O
bc	O
—	O
collect	O
all	O
contacts	O

from	O
the	O
Android	B-System
device	O
and	O
SIM	O
card	O
setForward	O
—	O
currently	O
not	O
implemented	O
,	O
but	O
can	O
be	O
used	O
to	O
hijack	O
the	O
infected	O
device	O
getForward	O
—	O
currently	O
not	O
implemented	O
,	O
but	O
can	O
be	O
used	O
to	O
hijack	O
the	O
infected	O
device	O
hasPkg	O
—	O
check	O
the	O
device	O
whether	O
a	O
specified	O
app	O
is	O
installed	O
or	O
not	O
setRingerMode	O
—	O
set	O
the	O
device	O
’	O
s	O
ringer	O
mode	O
setRecEnable	O
—	O
set	O
the	O
device	O
’	O
s	O
ringer	O
mode	O
as	O
silent	O
reqState	O
—	O
get	O
a	O
detailed	O
phone	O
connection	O
status	O
,	O
which	O
includes	O
activated	O
network	O
and	O
Wi-Fi	O
(	O
with	O
or	O
without	O
password	O
)	O
showHome	O
—	O

force	O
the	O
device	O
’	O
s	O
back	O
to	O
the	O
home	O
screen	O
getnpki	O
:	O
get	O
files/content	O
from	O
the	O
folder	O
named	O
NPKI	O
(	O
contains	O
certificates	O
related	O
to	O
financial	O
transactions	O
)	O
http	O
—	O
access	O
a	O
specified	O
network	O
using	O
HttpURLConnection	O
onRecordAction	O
—	O
simulate	O
a	O
number-dialed	O
tone	O
call	O
—	O
call	O
a	O
specified	O
number	O
get_apps	O
—	O
get	O
all	O
the	O
apps	O
installed	O
on	O
the	O
device	O
show_fs_float_window	O
—	O
show	O
a	O
full-screen	O
window	O
for	O
phishing	O
Of	O
note	O
is	O
XLoader	B-Malware
’	O
s	O
abuse	O
of	O
the	O
WebSocket	O
protocol	O
(	O
supported	O
in	O
many	O
browsers	O

and	O
web	O
applications	O
)	O
via	O
ws	O
(	O
WebSockets	O
)	O
or	O
wss	O
(	O
WebSockets	O
over	O
SSL/TLS	O
)	O
to	O
communicate	O
with	O
its	O
C	O
&	O
C	O
servers	O
.	O

The	O
URLs	O
—	O
abused	O
as	O
part	O
of	O
XLoader	B-Malware
’	O
s	O
C	O
&	O
C	O
—	O
are	O
hidden	O
in	O
three	O
webpages	O
,	O
and	O
the	O
C	O
&	O
C	O
server	O
that	O
XLoader	B-Malware
connects	O
to	O
differ	O
per	O
region	O
.	O

The	O
abuse	O
of	O
the	O
WebSocket	O
protocol	O
provides	O
XLoader	B-Malware
with	O
a	O
persistent	O
connection	O
between	O
clients	O
and	O
servers	O
where	O
data	O
can	O
be	O
transported	O
any	O
time	O
.	O

XLoader	B-Malware
abuses	O
the	O
MessagePack	O
(	O
a	O
data	O
interchange	O
format	O
)	O
to	O
package	O
the	O
stolen	O
data	O
and	O
exfiltrate	O
it	O
via	O
the	O
WebSocket	O
protocol	O
for	O
faster	O
and	O
more	O
efficient	O
transmission	O
.	O

Mitigations	O
XLoader	B-Malware
will	O
not	O
download	O
malicious	O
apps	O
if	O
the	O
Android	O
device	O
uses	O
a	O
mobile	O
data	O
connection	O
.	O

Nevertheless	O
,	O
users	O
should	O
practice	O
proper	O
security	O
hygiene	O
to	O
mitigate	O
threats	O
that	O
may	O
take	O
advantage	O
of	O
a	O
home	O
or	O
business	O
router	O
’	O
s	O
security	O
gaps	O
.	O

Employ	O
stronger	O
credentials	O
,	O
for	O
instance	O
,	O
to	O
make	O
them	O
less	O
susceptible	O
to	O
unauthorized	O
access	O
.	O

Regularly	O
update	O
and	O
patch	O
the	O
router	O
’	O
s	O
software	O
and	O
firmware	O
to	O
prevent	O
exploits	O
,	O
and	O
enable	O
its	O
built-in	O
firewall	O
.	O

For	O
system	O
administrators	O
and	O
information	O
security	O
professionals	O
,	O
configuring	O
the	O
router	O
to	O
be	O
more	O
resistant	O
to	O
attacks	O
like	O
DNS	O
cache	O
poisoning	O
can	O
help	O
mitigate	O
similar	O
threats	O
.	O

Everyday	O
users	O
can	O
do	O
the	O
same	O
by	O
checking	O
the	O
router	O
’	O
s	O
DNS	O
settings	O
if	O
they	O
’	O
ve	O
been	O
modified	O
.	O

Even	O
threats	O
like	O
DNS	O
cache	O
poisoning	O
employ	O
social	O
engineering	O
,	O
so	O
users	O
should	O
also	O
be	O
more	O
prudent	O
against	O
suspicious	O
or	O
unknown	O
messages	O
that	O
have	O
telltale	O
signs	O
of	O
malware	O
.	O

We	O
have	O
worked	O
with	O
Google	B-Organization
and	O
they	O
ensure	O
that	O
Google	B-System
Play	I-System
Protect	I-System
proactively	O
catches	O
apps	O
of	O
this	O
nature	O
.	O

No	O
instances	O
of	O
these	O
apps	O
were	O
found	O
in	O
Google	B-System
Play	I-System
.	O

September	O
08	O
,	O
2020	O
TikTok	B-System
Spyware	O
A	O
detailed	O
analysis	O
of	O
spyware	O
masquerading	O
as	O
TikTok	B-System
A	O
recent	O
threat	O
to	O
ban	O
TikTok	B-System
in	O
the	O
United	O
States	O
has	O
taken	O
the	O
internet	O
by	O
storm	O
and	O
received	O
mixed	O
reactions	O
from	O
social	O
media	O
and	O
internet	O
users	O
.	O

U.S.	O
President	O
Donald	O
Trump	O
has	O
ordered	O
ByteDance	B-Organization
,	O
the	O
parent	O
company	O
of	O
TikTok	B-System
,	O
to	O
sell	O
its	O
U.S.	O
TikTok	B-System
assets	O
and	O
also	O
issued	O
executive	O
orders	O
that	O
would	O
ban	O
the	O
social	O
media	O
apps	O
TikTok	B-System
and	O
WeChat	B-System
from	O
operating	O
in	O
the	O
U.S.	O
if	O
the	O
sale	O
doesn	O
’	O
t	O
happen	O
in	O
the	O
next	O
few	O
weeks	O
.	O

On	O
the	O
other	O
side	O
,	O
ByteDance	B-Organization
has	O
filed	O
a	O
lawsuit	O
suing	O
the	O
Trump	O
administration	O
.	O

When	O
popular	O
applications	O
come	O
under	O
fire	O
and	O
are	O
featured	O
prominently	O
in	O
the	O
news	O
,	O
hackers	O
get	O
excited	O
as	O
these	O
newsworthy	O
apps	O
can	O
become	O
their	O
latest	O
target	O
.	O

And	O
TikTok	B-System
is	O
no	O
exception	O
.	O

Generally	O
,	O
after	O
an	O
application	O
gets	O
banned	O
from	O
an	O
official	O
app	O
store	O
,	O
such	O
as	O
Google	B-System
Play	I-System
,	O
users	O
try	O
to	O
find	O
alternative	O
ways	O
to	O
download	O
the	O
app	O
.	O

In	O
doing	O
so	O
,	O
users	O
can	O
become	O
victims	O
to	O
malicious	O
apps	O
portraying	O
themselves	O
as	O
the	O
original	O
app	O
.	O

Recently	O
there	O
was	O
a	O
huge	O
wave	O
of	O
SMS	O
messages	O
,	O
as	O
well	O
as	O
Whatsapp	B-System
messages	O
,	O
making	O
the	O
rounds	O
asking	O
users	O
to	O
download	O
the	O
latest	O
version	O
of	O
TikTok	B-System
at	O
hxxp	B-Indicator
:	I-Indicator
//tiny	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
cc/TiktokPro	I-Indicator
.	O

In	O
reality	O
,	O
this	O
downloaded	O
app	O
is	O
a	O
fake	O
app	O
that	O
asks	O
for	O
credentials	O
and	O
Android	B-System
permissions	O
(	O
including	O
camera	O
and	O
phone	O
permissions	O
)	O
,	O
resulting	O
in	O
the	O
user	O
being	O
bombarded	O
with	O
advertisements	O
.	O

Recently	O
,	O
we	O
have	O
come	O
across	O
another	O
variant	O
of	O
this	O
app	O
portraying	O
itself	O
as	O
TikTok	B-System
Pro	I-System
,	O
but	O
this	O
is	O
a	O
full-fledged	O
spyware	O
with	O
premium	O
features	O
to	O
spy	O
on	O
victim	O
with	O
ease	O
.	O

(	O
Please	O
note	O
this	O
is	O
a	O
different	O
app	O
and	O
not	O
the	O
same	O
as	O
the	O
one	O
being	O
spread	O
by	O
hxxp	B-Indicator
:	I-Indicator
//tiny	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
cc/TiktokPro	I-Indicator
.	I-Indicator

)	I-Indicator
Technical	O
Analysis	O
App	O
Name	O
:	O
TikTok	B-System
Pro	I-System
Hash	O
:	O
9fed52ee7312e217bd10d6a156c8b988	B-Indicator
Package	O
Name	O
:	O
com.example.dat.a8andoserverx	B-Indicator
Upon	O
installation	O
,	O
the	O
spyware	O
portrays	O
itself	O
as	O
TikTok	B-System
using	O
the	O
name	O
TikTok	B-System
Pro	I-System
.	O

As	O
soon	O
as	O
a	O
user	O
tries	O
to	O
open	O
the	O
app	O
,	O
it	O
launches	O
a	O
fake	O
notification	O
and	O
soon	O
the	O
notification	O
as	O
well	O
as	O
the	O
app	O
icon	O
disappears	O
.	O

This	O
fake	O
notification	O
tactic	O
is	O
used	O
to	O
redirect	O
the	O
user	O
's	O
attention	O
,	O
meanwhile	O
the	O
app	O
hides	O
itself	O
,	O
making	O
the	O
user	O
believe	O
the	O
app	O
to	O
be	O
faulty	O
.	O

This	O
functionality	O
can	O
be	O
seen	O
in	O
Figure	O
1	O
.	O

App	O
Icon	O
Figure	O
1	O
:	O
App	O
icon	O
and	O
fake	O
notification	O
.	O

Behind	O
the	O
scenes	O
,	O
there	O
are	O
number	O
of	O
process	O
occurring	O
simultaneously	O
.	O

First	O
,	O
an	O
activity	O
named	O
MainActivity	O
fires	O
up	O
,	O
taking	O
care	O
of	O
hiding	O
the	O
icon	O
and	O
showing	O
the	O
fake	O
notification	O
.	O

It	O
also	O
starts	O
an	O
Android	B-System
service	O
named	O
MainService	O
.	O

The	O
spyware	O
also	O
appears	O
to	O
have	O
an	O
additional	O
payload	O
stored	O
under	O
the	O
/res/raw/	O
directory	O
.	O

This	O
is	O
a	O
common	O
technique	O
used	O
by	O
malware	O
developers	O
to	O
bundle	O
the	O
main	O
payload	O
inside	O
the	O
Android	B-System
package	O
to	O
avoid	O
easy	O
detection	O
.	O

As	O
seen	O
in	O
Figure	O
2	O
,	O
the	O
app	O
tries	O
to	O
open	O
the	O
payload	O
from	O
the	O
/res/raw/	O
directory	O
and	O
generate	O
an	O
additional	O
Android	B-System
Package	I-System
Kit	I-System
(	O
APK	O
)	O
named	O
.app.apk	B-Indicator
:	O
Decoy	O
Code	O
Figure	O
2	O
:	O
The	O
decoy	O
code	O
for	O
the	O
fake	O
TikTok	B-System
.	O

Upon	O
analysis	O
,	O
we	O
discovered	O
that	O
this	O
is	O
a	O
decoy	O
functionality	O
and	O
no	O
new	O
payload	O
is	O
generated	O
.	O

The	O
conditions	O
to	O
build	O
an	O
additional	O
payload	O
are	O
never	O
met	O
.	O

Going	O
one	O
step	O
further	O
,	O
we	O
rebuilt	O
the	O
malware	O
to	O
execute	O
the	O
apparent	O
functionality	O
of	O
generating	O
a	O
payload	O
,	O
but	O
discovered	O
that	O
the	O
APK	O
stored	O
in	O
the	O
/res/raw/	O
directory	O
is	O
empty	O
.	O

The	O
placement	O
of	O
the	O
decoy	O
functionality	O
is	O
likely	O
designed	O
to	O
confuse	O
the	O
malware	O
researchers	O
.	O

It	O
is	O
also	O
possible	O
that	O
this	O
functionality	O
is	O
under	O
development	O
,	O
making	O
this	O
placeholder	O
code	O
incomplete	O
.	O

Coming	O
back	O
to	O
the	O
execution	O
flow	O
,	O
once	O
the	O
spyware	O
hides	O
itself	O
,	O
it	O
starts	O
an	O
Android	B-System
service	O
named	O
MainService	O
.	O

Android	B-System
services	O
are	O
components	O
that	O
can	O
be	O
made	O
to	O
execute	O
independently	O
in	O
the	O
background	O
without	O
the	O
victim	O
's	O
knowledge	O
.	O

MainService	O
is	O
the	O
brain	O
of	O
this	O
spyware	O
and	O
controls	O
almost	O
everything—from	O
stealing	O
the	O
victim	O
's	O
data	O
to	O
deleting	O
it	O
.	O

All	O
of	O
its	O
capabilities	O
are	O
discussed	O
later	O
in	O
this	O
blog	O
.	O

Hide	O
Icon	O
Figure	O
3	O
:	O
Code	O
showing	O
the	O
hiding	O
icon	O
and	O
starting	O
service	O
.	O

As	O
MainService	O
is	O
the	O
main	O
controller	O
,	O
the	O
developer	O
has	O
taken	O
the	O
appropriate	O
actions	O
to	O
keep	O
it	O
functional	O
and	O
running	O
at	O
all	O
times	O
.	O

The	O
malware	O
developer	O
uses	O
various	O
tactics	O
to	O
do	O
so	O
,	O
and	O
one	O
of	O
them	O
is	O
using	O
Android	B-System
's	O
broadcast	O
receivers	O
.	O

Broadcast	O
receivers	O
are	O
components	O
that	O
allow	O
you	O
to	O
register	O
for	O
various	O
Android	B-System
events	O
.	O

In	O
this	O
case	O
,	O
it	O
registers	O
three	O
broadcast	O
receivers	O
:	O
MyReceiver	O
-	O
Triggers	O
when	O
the	O
device	O
is	O
booted	O
.	O

Intercept	O
Call	O
-	O
Triggers	O
on	O
incoming	O
and	O
outgoing	O
calls	O
.	O

AlarmReceiver	O
-	O
Triggers	O
every	O
three	O
minutes	O
.	O

MyReceiver	O
and	O
AlarmReceiver	O
start	O
the	O
MainService	O
whenever	O
appropriate	O
events	O
occur	O
.	O

This	O
tactic	O
is	O
very	O
common	O
among	O
malware	O
developers	O
to	O
ensure	O
the	O
malware	O
is	O
not	O
killed	O
by	O
the	O
Android	B-System
OS	O
or	O
by	O
any	O
other	O
means	O
.	O

Figure	O
4	O
shows	O
MyReceiver	O
in	O
action	O
where	O
it	O
eventually	O
calls	O
the	O
MainService	O
service	O
.	O

Broadcast	O
Receiver	O
Figure	O
4	O
:	O
MyReceiver	O
broadcast	O
receiver	O
.	O

The	O
InterceptCall	O
receiver	O
is	O
triggered	O
whenever	O
there	O
is	O
an	O
incoming	O
or	O
outgoing	O
call	O
.	O

It	O
sets	O
particular	O
parameters	O
in	O
relation	O
to	O
call	O
details	O
and	O
a	O
further	O
service	O
named	O
calls	O
takes	O
the	O
control	O
as	O
seen	O
in	O
Figure	O
5	O
.	O

Call	O
Service	O
Figure	O
5	O
:	O
Code	O
for	O
the	O
calls	O
service	O
As	O
seen	O
above	O
,	O
the	O
calls	O
service	O
stores	O
incoming	O
call	O
details	O
in	O
.mp3	O
format	O
in	O
the	O
/sdcard/DCIM/.dat/	O
directory	O
with	O
file	O
name	O
appended	O
with	O
"	O
In_	O
''	O
for	O
incoming	O
calls	O
and	O
"	O
Out_	O
''	O
for	O
outgoing	O
calls	O
.	O

How	O
these	O
recorded	O
calls	O
are	O
sent	O
to	O
the	O
command	O
and	O
control	O
server	O
(	O
CnC	O
)	O
is	O
taken	O
care	O
of	O
by	O
MainService	O
,	O
which	O
is	O
discussed	O
next	O
.	O

MainService	O
is	O
the	O
central	O
controller	O
of	O
this	O
spyware	O
.	O

It	O
controls	O
each	O
and	O
every	O
functionality	O
based	O
on	O
the	O
commands	O
sent	O
by	O
the	O
command	O
and	O
control	O
(	O
C	O
&	O
C	O
)	O
server	O
.	O

As	O
soon	O
as	O
this	O
service	O
is	O
started	O
,	O
it	O
creates	O
two	O
processes	O
that	O
take	O
care	O
of	O
connection	O
and	O
disconnection	O
to	O
the	O
C	O
&	O
C	O
server	O
.	O

This	O
functionality	O
can	O
be	O
seen	O
in	O
Figure	O
6	O
.	O

TimerTask	O
Figure	O
6	O
:	O
The	O
timer	O
task	O
.	O

MainService	O
has	O
the	O
following	O
capabilities	O
:	O
Steal	O
SMS	O
messages	O
Send	O
SMS	O
messages	O
Steal	O
the	O
victim	O
's	O
location	O
Capture	O
photos	O
Execute	O
commands	O
Capture	O
screenshots	O
Call	O
phone	O
numbers	O
Initiate	O
other	O
apps	O
Steal	O
Facebook	B-System
credentials	O
,	O
etc	O
All	O
of	O
the	O
above	O
functionalities	O
take	O
place	O
on	O
the	O
basis	O
of	O
commands	O
sent	O
by	O
the	O
attacker	O
.	O

Stolen	O
data	O
is	O
stored	O
in	O
external	O
storage	O
under	O
the	O
/DCIM/	O
directory	O
with	O
a	O
hidden	O
sub-directory	O
named	O
"	O
.dat	O
''	O
.	O

Below	O
is	O
the	O
list	O
of	O
all	O
the	O
commands	O
catered	O
by	O
the	O
C	O
&	O
C	O
server	O
.	O

Command	O
Action	O
Unistxcr	O
Restart	O
the	O
app	O
dowsizetr	O
Send	O
the	O
file	O
stored	O
in	O
the	O
/sdcard/DCIM/.dat/	O
directory	O
to	O
the	O
C	O
&	O
C	O
server	O
Caspylistx	O
Get	O
a	O
list	O
of	O
all	O
hidden	O
files	O
in	O
the	O
/DCIM/.dat/	O
directory	O
spxcheck	O
Check	O
whether	O
call	O
details	O
are	O
collected	O
by	O
the	O
spyware	O
S8p8y0	O
Delete	O
call	O
details	O
stored	O
by	O
the	O
spyware	O
screXmex	O
Take	O
screenshots	O
of	O
the	O
device	O
screen	O
Batrxiops	O
Check	O
battery	O
status	O
L4oclOCMAWS	O
Fetch	O
the	O
victim	O
's	O
location	O
GUIFXB	O
Launch	O

the	O
fake	O
Facebook	B-System
login	O
page	O
IODBSSUEEZ	O
Send	O
a	O
file	O
containing	O
stolen	O
Facebook	B-System
credentials	O
to	O
the	O
C	O
&	O
C	O
server	O
FdelSRRT	O
Delete	O
files	O
containing	O
stolen	O
Facebook	B-System
credentials	O
chkstzeaw	O
Launch	O
Facebook	B-System
LUNAPXER	O
Launch	O
apps	O
according	O
to	O
the	O
package	O
name	O
sent	O
by	O
the	O
C	O
&	O
C	O
server	O
Gapxplister	O
Get	O
a	O
list	O
of	O
all	O
installed	O
applications	O
DOTRall8xxe	O
Zip	O
all	O
the	O
stolen	O
files	O
and	O
store	O
in	O
the	O
/DCIM/.dat/	O
directory	O
Acouxacour	O
Get	O
a	O
list	O
of	O
accounts	O
on	O
the	O
victim	O
's	O
device	O
Fimxmiisx	O
Open	O
the	O
camera	O

Scxreexcv4	O
Capture	O
an	O
image	O
micmokmi8x	O
Capture	O
audio	O
Yufsssp	O
Get	O
latitude	O
and	O
longitude	O
GExCaalsss7	O
Get	O
call	O
logs	O
PHOCAs7	O
Call	O
phone	O
numbers	O
sent	O
by	O
the	O
C	O
&	O
C	O
server	O
Gxextsxms	O
Get	O
a	O
list	O
of	O
inbox	O
SMS	O
messages	O
Msppossag	O
Send	O
SMS	O
with	O
message	O
body	O
sent	O
by	O
the	O
C	O
&	O
C	O
server	O
Getconstactx	O
Get	O
a	O
list	O
of	O
all	O
contacts	O
Rinxgosa	O
Play	O
a	O
ringtone	O
bithsssp64	O
Execute	O
commands	O
sent	O
by	O
the	O
C	O
&	O
C	O
server	O
DOWdeletx	O
Deletes	O

the	O
file	O
specified	O
by	O
the	O
C	O
&	O
C	O
server	O
Deldatall8	O
Delete	O
all	O
files	O
stored	O
in	O
the	O
/sdcard/DCIM/.dat/	O
directory	O
We	O
do	O
n't	O
have	O
the	O
space	O
to	O
cover	O
all	O
of	O
the	O
commands	O
,	O
but	O
let	O
's	O
take	O
a	O
look	O
at	O
some	O
of	O
the	O
major	O
ones	O
.	O

Facebook	B-System
phishing	O
One	O
of	O
the	O
interesting	O
features	O
of	O
this	O
spyware	O
is	O
the	O
ability	O
to	O
steal	O
Facebook	B-System
credentials	O
using	O
a	O
fake	O
login	O
page	O
,	O
similar	O
to	O
phishing	O
.	O

Upon	O
receiving	O
the	O
command	O
GUIFXB	O
,	O
the	O
spyware	O
launches	O
a	O
fake	O
Facebook	B-System
login	O
page	O
.	O

As	O
soon	O
as	O
the	O
victim	O
tries	O
to	O
log	O
in	O
,	O
it	O
stores	O
the	O
victim	O
's	O
credentials	O
in	O
/storage/0/DCIM/.fdat	O
Facebook	B-System
Login	O
Figure	O
7	O
:	O
Fake	O
Facebook	B-System
login	O
The	O
second	O
command	O
is	O
IODBSSUEEZ	O
,	O
which	O
further	O
sends	O
stolen	O
credentials	O
to	O
the	O
C	O
&	O
C	O
server	O
,	O
as	O
seen	O
in	O
Figure	O
8	O
.	O

Stolen	O
Data	O
Figure	O
8	O
:	O
Sending	O
data	O
to	O
the	O
attacker	O
.	O

This	O
functionality	O
can	O
be	O
easily	O
further	O
extended	O
to	O
steal	O
other	O
information	O
,	O
such	O
as	O
bank	O
credentials	O
,	O
although	O
we	O
did	O
not	O
see	O
any	O
banks	O
being	O
targeted	O
in	O
this	O
attack	O
.	O

Calling	O
functionality	O
Command	O
PHOCAs7	O
initiates	O
calling	O
functionality	O
.	O

The	O
number	O
to	O
call	O
is	O
received	O
along	O
with	O
the	O
command	O
,	O
as	O
seen	O
in	O
Figure	O
9	O
.	O

Call	O
Command	O
Figure	O
9	O
:	O
The	O
calling	O
functionality	O
.	O

The	O
phone	O
number	O
is	O
fetched	O
from	O
a	O
response	O
from	O
the	O
C	O
&	O
C	O
server	O
and	O
is	O
stored	O
in	O
str3	O
variable	O
,	O
which	O
further	O
is	O
utilized	O
using	O
the	O
tel	O
:	O
function	O
.	O

Stealing	O
SMS	O
The	O
Gxextsxms	O
command	O
is	O
responsible	O
for	O
fetching	O
all	O
the	O
SMS	O
messages	O
from	O
the	O
victim	O
's	O
device	O
and	O
sending	O
it	O
over	O
to	O
the	O
C	O
&	O
C	O
server	O
.	O

Stealing	O
SMS	O
Figure	O
10	O
:	O
Stealing	O
SMS	O
messages	O
.	O

Similarly	O
,	O
there	O
are	O
many	O
crucial	O
commands	O
that	O
further	O
allow	O
this	O
spyware	O
to	O
perform	O
additional	O
functionality	O
,	O
such	O
as	O
executing	O
commands	O
sent	O
by	O
the	O
C	O
&	O
C	O
,	O
clicking	O
photos	O
,	O
capturing	O
screenshots	O
,	O
stealing	O
location	O
information	O
,	O
and	O
more	O
.	O

Further	O
analysis	O
Upon	O
further	O
research	O
,	O
we	O
found	O
this	O
spyware	O
to	O
be	O
developed	O
by	O
a	O
framework	O
similar	O
to	O
Spynote	B-Malware
and	O
Spymax	B-Malware
,	O
meaning	O
this	O
could	O
be	O
an	O
updated	O
version	O
of	O
these	O
Trojan	O
builders	O
,	O
which	O
allow	O
anyone	O
,	O
even	O
with	O
limited	O
knowledge	O
,	O
to	O
develop	O
full-fledged	O
spyware	O
.	O

Many	O
of	O
the	O
functionalities	O
seen	O
in	O
this	O
spyware	O
are	O
similar	O
to	O
Spynote	B-Malware
and	O
Spymax	B-Malware
based	O
on	O
the	O
samples	O
we	O
analyzed	O
with	O
some	O
modifications	O
.	O

This	O
spyware	O
sample	O
communicates	O
over	O
dynamic	O
DNS	O
.	O

By	O
doing	O
so	O
,	O
attackers	O
can	O
easily	O
set	O
up	O
the	O
Trojan	O
to	O
communicate	O
back	O
to	O
them	O
without	O
any	O
need	O
for	O
high-end	O
servers	O
.	O

Other	O
common	O
functionalities	O
include	O
executing	O
commands	O
received	O
from	O
the	O
attacker	O
,	O
taking	O
screenshots	O
of	O
the	O
victim	O
's	O
device	O
,	O
fetching	O
locations	O
,	O
stealing	O
SMS	O
messages	O
and	O
most	O
common	O
features	O
that	O
every	O
spyware	O
may	O
poses	O
.	O

Stealing	O
Facebook	B-Organization
credentials	O
using	O
fake	O
Facebook	B-Organization
activity	O
is	O
something	O
we	O
did	O
n't	O
observe	O
in	O
Spynote/Spymax	B-Malware
versions	O
but	O
was	O
seen	O
in	O
this	O
spyware	O
.	O

This	O
framework	O
allows	O
anyone	O
to	O
develop	O
a	O
malicious	O
app	O
with	O
the	O
desired	O
icon	O
and	O
communication	O
address	O
.	O

Some	O
of	O
the	O
icons	O
used	O
can	O
be	O
seen	O
below	O
.	O

We	O
found	O
280	O
such	O
apps	O
in	O
the	O
past	O
three	O
months	O
.	O

A	O
complete	O
list	O
of	O
hashes	O
can	O
be	O
found	O
here	O
.	O

icons	O
Figure	O
11	O
:	O
Icons	O
used	O
to	O
pose	O
as	O
famous	O
apps	O
.	O

All	O
of	O
these	O
apps	O
are	O
developed	O
by	O
the	O
same	O
framework	O
and	O
hence	O
have	O
the	O
same	O
package	O
name	O
and	O
certificate	O
information	O
as	O
seen	O
in	O
Figure	O
12.	O
certificate	O
Figure	O
12	O
:	O
Package	O
name	O
and	O
certificate	O
information	O
.	O

Conclusion	O
Due	O
to	O
the	O
ubiquitous	O
nature	O
of	O
mobile	O
devices	O
and	O
the	O
widespread	O
use	O
of	O
Android	B-System
,	O
it	O
is	O
very	O
easy	O
for	O
attackers	O
to	O
victimize	O
Android	B-System
users	O
.	O

In	O
such	O
situations	O
,	O
mobile	O
users	O
should	O
always	O
take	O
the	O
utmost	O
precautions	O
while	O
downloading	O
any	O
applications	O
from	O
the	O
internet	O
.	O

It	O
is	O
very	O
easy	O
to	O
trick	O
victims	O
to	O
fall	O
for	O
such	O
attacks	O
.	O

Users	O
looking	O
forward	O
to	O
using	O
the	O
TikTok	B-System
app	O
amidst	O
the	O
ban	O
might	O
look	O
for	O
alternative	O
methods	O
to	O
download	O
the	O
app	O
.	O

In	O
doing	O
so	O
,	O
users	O
can	O
mistakenly	O
install	O
malicious	O
apps	O
,	O
such	O
as	O
the	O
spyware	O
mentioned	O
in	O
this	O
blog	O
.	O

The	O
precautions	O
you	O
take	O
online	O
have	O
been	O
covered	O
extensively	O
in	O
almost	O
all	O
of	O
our	O
blogs	O
;	O
even	O
so	O
,	O
we	O
believe	O
this	O
information	O
bears	O
repeating	O
.	O

Please	O
follow	O
these	O
basic	O
precautions	O
during	O
the	O
current	O
crisis—and	O
at	O
all	O
times	O
:	O
Install	O
apps	O
only	O
from	O
official	O
stores	O
,	O
such	O
as	O
Google	B-System
Play	I-System
.	O

Never	O
click	O
on	O
unknown	O
links	O
received	O
through	O
ads	O
,	O
SMS	O
messages	O
,	O
emails	O
,	O
or	O
the	O
like	O
.	O

Always	O
keep	O
the	O
"	O
Unknown	O
Sources	O
''	O
option	O
disabled	O
in	O
the	O
Android	B-System
device	O
.	O

This	O
disallows	O
apps	O
to	O
be	O
installed	O
on	O
your	O
device	O
from	O
unknown	O
sources	O
.	O

We	O
would	O
also	O
like	O
to	O
mention	O
that	O
if	O
you	O
come	O
across	O
an	O
app	O
hiding	O
it	O
's	O
icon	O
,	O
always	O
try	O
to	O
search	O
for	O
the	O
app	O
in	O
your	O
device	O
settings	O
(	O
by	O
going	O
to	O
Settings	O
-	O
>	O
Apps	O
-	O
>	O
Search	O
for	O
icon	O
that	O
was	O
hidden	O
)	O
.	O

In	O
the	O
case	O
of	O
this	O
spyware	O
,	O
search	O
for	O
app	O
named	O
TikTok	B-System
Pro	I-System
.	O

MITRE	B-Organization
TAGS	O
Action	O
Tag	O
ID	O
App	O
auto-start	O
at	O
device	O
boot	O
T1402	O
Input	O
prompt	O
T1411	O
Capture	O
SMS	O
messages	O
T1412	O
Application	O
discovery	O
T1418	O
Capture	O
audio	O
T1429	O
Location	O
tracking	O
T1430	O
Access	O
contact	O
list	O
T1432	O
Access	O
call	O
log	O
T1433	O
Commonly	O
used	O
port	O
T1436	O
Standard	O
application	O
layer	O
protocol	O
T1437	O
Masquerage	O
as	O
legitimate	O
application	O
T1444	O
Suppress	O
application	O
icon	O
T1508	O
Capture	O
camera	O
T1512	O
Screen	O
capture	O
T1513	O
Foreground	O
persistence	O
T1541	O
DualToy	B-Malware
:	O
New	O
Windows	B-System
Trojan	O
Sideloads	O
Risky	O
Apps	O
to	O
Android	B-System
and	O
iOS	B-System
Devices	O

By	O
Claud	O
Xiao	O
September	O
13	O
,	O
2016	O
at	O
5:00	O
AM	O
Over	O
the	O
past	O
two	O
years	O
,	O
we	O
’	O
ve	O
observed	O
many	O
cases	O
of	O
Microsoft	B-System
Windows	I-System
and	O
Apple	B-System
iOS	I-System
malware	O
designed	O
to	O
attack	O
mobile	O
devices	O
.	O

This	O
attack	O
vector	O
is	O
increasingly	O
popular	O
with	O
malicious	O
actors	O
as	O
almost	O
everyone	O
on	O
the	O
planet	O
carries	O
at	O
least	O
one	O
mobile	O
device	O
they	O
interact	O
with	O
throughout	O
any	O
given	O
day	O
.	O

Thanks	O
to	O
a	O
relative	O
lack	O
of	O
security	O
controls	O
applied	O
to	O
mobile	O
devices	O
,	O
these	O
devices	O
have	O
become	O
very	O
attractive	O
targets	O
for	O
a	O
broad	O
range	O
of	O
malicious	O
actors	O
.	O

For	O
example	O
:	O
WireLurker	B-Malware
installed	O
malicious	O
apps	O
on	O
non-jailbroken	O
iPhones	O
Six	O
different	O
Trojan	O
,	O
Adware	O
and	O
HackTool	B-Malware
families	I-Malware
launched	O
“	O
BackStab	O
”	O
attacks	O
to	O
steal	O
backup	O
archives	O
of	O
iOS	B-System
and	O
BlackBerry	B-System
devices	O
The	O
HackingTeam	B-Malware
’	O
s	O
RCS	B-Malware
delivered	O
its	O
Spyware	O
from	O
infected	O
PCs	O
and	O
Macs	O
to	O
jailbroken	O
iOS	B-System
devices	O
and	O
BlackBerry	B-System
phones	O
Recently	O
,	O
we	O
discovered	O
another	O
Windows	B-System
Trojan	O
we	O
named	O
“	O
DualToy	B-Malware
”	O
which	O
side	O
loads	O
malicious	O
or	O
risky	O
apps	O
to	O
both	O
Android	B-System
and	O
iOS	B-System
devices	O
via	O
a	O
USB	B-System
connection	O
.	O

When	O
DualToy	B-Malware
began	O
to	O
spread	O
in	O
January	O
2015	O
,	O
it	O
was	O
only	O
capable	O
of	O
infecting	O
Android	B-System
devices	O
.	O

However	O
,	O
within	O
six	O
months	O
the	O
malicious	O
actors	O
added	O
the	O
capability	O
to	O
infect	O
iOS	B-System
devices	O
.	O

DualToy	B-Malware
is	O
still	O
active	O
and	O
we	O
have	O
detected	O
over	O
8,000	O
unique	O
samples	O
belonging	O
to	O
this	O
Trojan	O
family	O
to	O
date	O
.	O

It	O
mainly	O
targets	O
Chinese	O
users	O
,	O
but	O
has	O
also	O
successfully	O
affected	O
people	O
and	O
organizations	O
in	O
the	O
United	O
States	O
,	O
United	O
Kingdom	O
,	O
Thailand	O
,	O
Spain	O
,	O
and	O
Ireland	O
.	O

Credential	O
phishing	O
and	O
an	O
Android	B-System
banking	O
Trojan	O
combine	O
in	O
Austrian	O
mobile	O
attacks	O
NOVEMBER	O
03	O
,	O
2017	O
Overview	O
Credential	O
phishing	O
,	O
banking	O
Trojans	O
,	O
and	O
credit	O
card	O
phishing	O
schemes	O
are	O
common	O
threats	O
that	O
we	O
regularly	O
observe	O
both	O
at	O
scale	O
and	O
in	O
more	O
targeted	O
attacks	O
.	O

However	O
,	O
Proofpoint	B-Organization
researchers	O
have	O
recently	O
observed	O
phishing	O
attacks	O
that	O
incorporate	O
all	O
of	O
these	O
elements	O
in	O
a	O
single	O
,	O
multistep	O
scheme	O
involving	O
the	O
Marcher	B-Malware
Android	O
banking	O
Trojan	O
targeting	O
customers	O
of	O
large	O
Austrian	O
banks	O
.	O

Attacks	O
involving	O
Marcher	B-Malware
have	O
become	O
increasingly	O
sophisticated	O
,	O
with	O
documented	O
cases	O
involving	O
multiple	O
attack	O
vectors	O
and	O
a	O
variety	O
of	O
targeted	O
financial	O
services	O
and	O
communication	O
platforms	O
[	O
1	O
]	O
[	O
2	O
]	O
.	O

In	O
this	O
case	O
,	O
a	O
threat	O
actor	O
has	O
been	O
targeting	O
customers	O
of	O
Bank	O
Austria	O
,	O
Raiffeisen	O
Meine	O
Bank	O
,	O
and	O
Sparkasse	O
since	O
at	O
least	O
January	O
2017	O
.	O

The	O
attacks	O
described	O
here	O
begin	O
with	O
a	O
banking	O
credential	O
phishing	O
scheme	O
,	O
followed	O
by	O
an	O
attempt	O
to	O
trick	O
the	O
victim	O
into	O
installing	O
Marcher	B-Malware
,	O
and	O
finally	O
with	O
attempts	O
to	O
steal	O
credit	O
card	O
information	O
by	O
the	O
banking	O
Trojan	O
itself	O
.	O

Analysis	O
Marcher	B-Malware
is	O
frequently	O
distributed	O
via	O
SMS	O
,	O
but	O
in	O
this	O
case	O
,	O
victims	O
are	O
presented	O
with	O
a	O
link	O
in	O
an	O
email	O
.	O

Oftentimes	O
,	O
the	O
emailed	O
link	O
is	O
a	O
bit.ly	O
shortened	O
link	O
,	O
used	O
to	O
potentially	O
evade	O
detection	O
.	O

The	O
link	O
leads	O
to	O
a	O
phishing	O
page	O
that	O
asks	O
for	O
banking	O
login	O
credentials	O
or	O
an	O
account	O
number	O
and	O
PIN	O
.	O

Figure	O
1	O
shows	O
one	O
such	O
landing	O
page	O
using	O
stolen	O
branding	O
from	O
Bank	O
Austria	O
.	O

Figure	O
1	O
:	O
Landing	O
page	O
for	O
phishing	O
scheme	O
asking	O
for	O
the	O
victim	O
’	O
s	O
signatory	O
number	O
and	O
PIN	O
using	O
stolen	O
branding	O
from	O
Bank	B-System
Austria	I-System
Because	O
the	O
actor	O
delivered	O
phishing	O
links	O
using	O
the	O
bit.ly	B-Indicator
URL	O
shortener	O
,	O
we	O
can	O
access	O
delivery	O
statistics	O
for	O
this	O
particular	O
campaign	O
.	O

The	O
link	O
resolves	O
to	O
a	O
URL	O
designed	O
to	O
appear	O
legitimate	O
,	O
with	O
a	O
canonical	O
domain	O
of	O
sicher97140	B-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
info	I-Indicator
including	O
the	O
“	O
bankaustria	O
”	O
brand	O
.	O

Figure	O
2	O
:	O
Bit.ly	B-Indicator
statistics	O
for	O
a	O
phishing	O
landing	O
page	O
targeting	O
Bank	B-System
Austria	I-System
customers	O
The	O
actor	O
appears	O
to	O
have	O
recently	O
begun	O
using	O
“	O
.top	O
”	O
top-level	O
domains	O
(	O
TLDs	O
)	O
for	O
their	O
phishing	O
landing	O
pages	O
and	O
have	O
implemented	O
a	O
consistent	O
naming	O
structure	O
as	O
shown	O
below	O
.	O

Earlier	O
this	O
year	O
,	O
the	O
actor	O
used	O
“	O
.pw	O
”	O
TLDs	O
while	O
the	O
Bank	B-System
Austria	I-System
scheme	O
highlighted	O
above	O
used	O
“	O
.info	O
”	O
.	O

Some	O
recent	O
campaigns	O
against	O
other	O
bank	O
customers	O
also	O
used	O
“	O
.gdn	O
”	O
TLDs	O
.	O

Other	O
attacks	O
on	O
Bank	B-System
Austria	I-System
customers	O
that	O
we	O
observed	O
resolved	O
to	O
the	O
following	O
.top	O
domains	O
:	O
Oct	O
23	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id8817062	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
23	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id8817461	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
23	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id8817465	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
23	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id8817466	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
23	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id8817469	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
17	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id58712	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
17	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id58717	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
17	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id58729	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
17	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id58729	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
17	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id87721	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
Oct	O
17	O
,	O
2017	O
hxxp	B-Indicator
:	I-Indicator
//online.bankaustria.at.id87726	I-Indicator
[	I-Indicator
.	I-Indicator

]	I-Indicator
top/	I-Indicator
These	O
permutations	O
of	O
TLDs	O
and	O
canonical	O
domains	O
incorporating	O
the	O
legitimate	O
domain	O
expected	O
by	O
the	O
targeted	O
banking	O
customers	O
exemplifies	O
recent	O
trends	O
in	O
social	O
engineering	O
by	O
threat	O
actors	O
.	O

Just	O
as	O
threat	O
actors	O
may	O
use	O
stolen	O
branding	O
in	O
their	O
email	O
lures	O
to	O
trick	O
potential	O
victims	O
,	O
they	O
reproduce	O
a	O
legitimate	O
domain	O
name	O
in	O
a	O
fraudulent	O
domain	O
that	O
is	O
not	O
controlled	O
by	O
the	O
bank	O
.	O

Once	O
the	O
victim	O
enters	O
their	O
account	O
information	O
on	O
the	O
landing	O
page	O
,	O
the	O
phishing	O
attack	O
then	O
requests	O
that	O
the	O
user	O
log	O
in	O
with	O
their	O
email	O
address	O
and	O
phone	O
number	O
.	O

Figure	O
3	O
:	O
Step	O
two	O
of	O
the	O
credential	O
phish	O
asking	O
for	O
the	O
victim	O
’	O
s	O
email	O
address	O
and	O
phone	O
number	O
Having	O
stolen	O
the	O
victim	O
’	O
s	O
account	O
and	O
personal	O
information	O
,	O
the	O
scammer	O
introduces	O
a	O
social	O
engineering	O
scheme	O
,	O
informing	O
users	O
that	O
they	O
currently	O
do	O
not	O
have	O
the	O
“	O
Bank	B-System
Austria	I-System
Security	I-System
App	I-System
”	O
installed	O
on	O
their	O
smartphone	O
and	O
must	O
download	O
it	O
to	O
proceed	O
.	O

Figure	O
4	O
shows	O
the	O
download	O
prompt	O
for	O
this	O
fake	O
app	O
;	O
an	O
English	O
translation	O
follows	O
.	O

Figure	O
4	O
:	O
Alert	O
prompting	O
the	O
victim	O
to	O
download	O
an	O
Android	B-System
banking	I-System
app	I-System
(	O
English	O
translation	O
below	O
)	O
,	O
with	O
stolen	O
branding	O
and	O
fraudulent	O
copy	O
*	O
*	O
*	O
Translation	O
*	O
*	O
*	O
Dear	O
Customer	O
,	O
The	O
system	O
has	O
detected	O
that	O
the	O
Bank	B-System
Austria	I-System
Security	I-System
App	I-System
is	O
not	O
installed	O
on	O
your	O
smartphone	O
.	O

Due	O
to	O
new	O
EU	B-Organization
money	O
laundering	O
guidelines	O
,	O
the	O
new	O
Bank	B-System
Austria	I-System
security	I-System
app	I-System
is	O
mandatory	O
for	O
all	O
customers	O
who	O
have	O
a	O
mobile	O
phone	O
number	O
in	O
our	O
system	O
.	O

Please	O
install	O
the	O
app	O
immediately	O
to	O
avoid	O
blocking	O
your	O
account	O
.	O

Follow	O
the	O
instructions	O
at	O
the	O
bottom	O
of	O
this	O
page	O
.	O

Why	O
you	O
need	O
the	O
Bank	B-System
Austria	I-System
Security	I-System
App	I-System
:	O
Due	O
to	O
outdated	O
technology	O
of	O
the	O
mobile	O
network	O
important	O
data	O
such	O
as	O
mTan	O
SMS	O
and	O
online	O
banking	O
connections	O
are	O
transmitted	O
unencrypted	O
.	O

Our	O
security	O
app	O
allows	O
us	O
to	O
transmit	O
this	O
sensitive	O
data	O
encrypted	O
to	O
you	O
,	O
thus	O
increasing	O
the	O
security	O
that	O
you	O
will	O
not	O
suffer	O
any	O
financial	O
loss	O
.	O

Step	O
1	O
:	O
Download	O
Bank	B-System
Austria	I-System
Security	I-System
App	I-System
Download	O
the	O
Bank	O
Austria	O
security	O
app	O
to	O
your	O
Android	O
device	O
.	O

To	O
do	O
this	O
,	O
open	O
the	O
displayed	O
link	O
on	O
your	O
mobile	O
phone	O
by	O
typing	O
in	O
the	O
URL	O
field	O
of	O
your	O
browser	O
or	O
scan	O
the	O
displayed	O
QR	O
code	O
.	O

*	O
*	O
*	O
End	O
translation	O
*	O
*	O
*	O
The	O
phishing	O
template	O
then	O
presents	O
additional	O
instructions	O
for	O
installing	O
the	O
fake	O
security	O
application	O
(	O
Figure	O
5	O
)	O
:	O
Figure	O
5	O
:	O
Additional	O
instructions	O
telling	O
the	O
victim	O
to	O
give	O
the	O
app	O
the	O
requested	O
permissions	O
(	O
English	O
translation	O
below	O
)	O
,	O
with	O
stolen	O
branding	O
and	O
fraudulent	O
copy	O
*	O
*	O
*	O
Translation	O
*	O
*	O
*	O
Step	O
2	O
:	O
Allow	O
installation	O
Open	O
your	O
device	O
's	O
settings	O
,	O
select	O
Security	O
or	O
Applications	O
(	O
depending	O
on	O
the	O
device	O
)	O
,	O
and	O
check	O
Unknown	O
sources	O
.	O

Step	O
3	O
:	O
Run	O
installation	O
Start	O
the	O
Bank	B-System
Austria	I-System
security	I-System
app	I-System
from	O
the	O
notifications	O
or	O
your	O
download	O
folder	O
,	O
tap	O
Install	O
.	O

After	O
successful	O
installation	O
,	O
tap	O
Open	O
and	O
enable	O
the	O
device	O
administrator	O
.	O

Finished	O
!	O

*	O
*	O
*	O
End	O
translation	O
*	O
*	O
*	O
Referring	O
again	O
to	O
bit.ly	B-Indicator
,	O
we	O
can	O
see	O
click	O
statistics	O
for	O
this	O
campaign	O
(	O
Figure	O
6	O
)	O
.	O

Figure	O
6	O
:	O
bit.ly	B-Indicator
statistics	O
for	O
the	O
fake	O
Bank	B-System
Austria	I-System
Android	I-System
app	I-System
download	O
link	O
From	O
this	O
small	O
sample	O
,	O
we	O
see	O
that	O
7	O
%	O
of	O
visitors	O
clicked	O
through	O
to	O
download	O
the	O
application	O
,	O
which	O
is	O
actually	O
a	O
version	O
of	O
the	O
Marcher	B-Malware
banking	I-Malware
Trojan	I-Malware
named	O
“	O
BankAustria.apk	B-Indicator
”	O
,	O
continuing	O
the	O
fraudulent	O
use	O
of	O
the	O
bank	O
’	O
s	O
branding	O
to	O
fool	O
potential	O
victims	O
.	O

This	O
sample	O
is	O
similar	O
to	O
those	O
presented	O
in	O
other	O
recent	O
Marcher	B-Malware
analyses	O
[	O
1	O
]	O
[	O
2	O
]	O
.	O

This	O
particular	O
application	O
is	O
signed	O
with	O
a	O
fake	O
certificate	O
:	O
Owner	O
:	O
CN=Unknown	O
,	O
OU=Unknown	O
,	O
O=Unknown	O
,	O
L=Unknown	O
,	O
ST=Unknown	O
,	O
C=Unknown	O
Issuer	O
CN=Unknown	O
,	O
OU=Unknown	O
,	O
O=Unknown	O
,	O
L=Unknown	O
,	O
ST=Unknown	O
,	O
C=Unknown	O
Serial	O
:	O
1c9157d7	O
Validity	O
:	O
11/02/2017	O
00:16:46	O
03/20/2045	O
00:16:46	O
MD5	O
Hash	O
:	O
A8:55:46:32:15	B-Indicator

:	I-Indicator
A9	I-Indicator
:	I-Indicator
D5:95	I-Indicator
:	I-Indicator
A9:91	I-Indicator
:	I-Indicator
C2:91:77:5D:30	I-Indicator
:	I-Indicator
F6	I-Indicator
SHA1	O
Hash	O
:	O
32:17	B-Indicator
:	I-Indicator
E9:7E:06	I-Indicator
:	I-Indicator
FE:5D:84	I-Indicator
:	I-Indicator
BE:7C:14:0C	I-Indicator
:	I-Indicator
C6:2B:12:85	I-Indicator
:	I-Indicator
E7:03:9A:5F	I-Indicator
The	O
app	O
requests	O
extensive	O
permissions	O
during	O
installation	O
that	O
enable	O
a	O
range	O
of	O
activities	O
supported	O
by	O
the	O
malware	O
.	O

Those	O
permission	O
shown	O
in	O
bold	O
below	O
are	O
the	O
most	O
problematic	O
:	O
Allows	O
an	O
application	O
to	O
write	O
to	O
external	O
storage	O
.	O

Allows	O
an	O
application	O
to	O
read	O
from	O
external	O
storage	O
.	O

Allows	O
an	O
application	O
to	O
use	O
SIP	O
service	O
.	O

Allows	O
an	O
application	O
to	O
collect	O
battery	O
statistics	O
Allows	O
an	O
app	O
to	O
access	O
precise	O
location	O
.	O

Allows	O
an	O
application	O
to	O
receive	O
SMS	O
messages	O
.	O

Allows	O
an	O
application	O
to	O
send	O
SMS	O
messages	O
.	O

Allows	O
an	O
application	O
to	O
read	O
SMS	O
messages	O
.	O

Allows	O
an	O
application	O
to	O
write	O
SMS	O
messages	O
.	O

Allows	O
an	O
application	O
to	O
initiate	O
a	O
phone	O
call	O
without	O
going	O
through	O
the	O
Dialer	O
user	O
interface	O
for	O
the	O
user	O
to	O
confirm	O
the	O
call	O
.	O

Allows	O
applications	O
to	O
access	O
information	O
about	O
networks	O
.	O

Allows	O
applications	O
to	O
open	O
network	O
sockets	O
.	O

Allows	O
an	O
application	O
to	O
read	O
the	O
user	O
's	O
contacts	O
data	O
.	O

Allows	O
an	O
application	O
to	O
read	O
or	O
write	O
the	O
system	O
settings	O
.	O

Allows	O
an	O
application	O
to	O
force	O
the	O
device	O
to	O
lock	O
Allows	O
applications	O
to	O
access	O
information	O
about	O
Wi-Fi	O
networks	O
.	O

Allows	O
applications	O
to	O
change	O
Wi-Fi	O
connectivity	O
state	O
.	O

Allows	O
applications	O
to	O
change	O
network	O
connectivity	O
state	O
.	O

Analysis	O
of	O
the	O
malware	O
shows	O
that	O
it	O
uses	O
the	O
common	O
string	O
obfuscation	O
of	O
character	O
replacement	O
(	O
Figure	O
7	O
)	O
:	O
Figure	O
7	O
:	O
Encoded	O
Marcher	B-Malware
Strings	O
Figure	O
8	O
:	O
Decoded	O
Marcher	B-Malware
Strings	O
As	O
noted	O
,	O
the	O
application	O
requests	O
extensive	O
permissions	O
during	O
installation	O
;	O
Figure	O
9	O
shows	O
the	O
request	O
to	O
act	O
as	O
device	O
administrator	O
,	O
a	O
particular	O
permission	O
that	O
should	O
very	O
rarely	O
be	O
granted	O
to	O
an	O
app	O
.	O

Figure	O
9	O
:	O
Prompt	O
for	O
application	O
permissions	O
upon	O
installation	O
Figures	O
10	O
and	O
11	O
show	O
the	O
other	O
permission	O
screens	O
for	O
the	O
app	O
:	O
Figure	O
10	O
Figure	O
10	O
:	O
Part	O
1	O
of	O
the	O
permission	O
screen	O
for	O
the	O
app	O
Figure	O
11	O
:	O
Part	O
2	O
of	O
the	O
permission	O
screen	O
for	O
the	O
app	O
Once	O
installed	O
the	O
app	O
will	O
place	O
a	O
legitimate	O
looking	O
icon	O
on	O
the	O
phone	O
’	O
s	O
home	O
screen	O
,	O
again	O
using	O
branding	O
stolen	O
from	O
the	O
bank	O
.	O

Figure	O
12	O
:	O
Fake	B-System
Bank	I-System
Austria	I-System
Security	I-System
application	I-System
icon	O
In	O
addition	O
to	O
operating	O
as	O
a	O
banking	O
Trojan	O
,	O
overlaying	O
a	O
legitimate	O
banking	O
app	O
with	O
an	O
indistinguishable	O
credential	O
theft	O
page	O
,	O
the	O
malware	O
also	O
asks	O
for	O
credit	O
card	O
information	O
from	O
the	O
user	O
when	O
they	O
open	O
applications	O
such	O
as	O
the	O
Google	B-System
Play	I-System
store	O
.	O

Figure	O
13	O
:	O
Popup	O
asking	O
for	O
a	O
credit	O
card	O
number	O
The	O
application	O
also	O
supports	O
stealing	O
credit	O
card	O
verification	O
information	O
(	O
Figures	O
14	O
and	O
15	O
)	O
.	O

Figure	O
14	O
:	O
Information	O
theft	O
via	O
fake	O
credit	O
card	O
verification	O
using	O
stolen	O
branding	O
Figure	O
15	O
:	O
Information	O
theft	O
via	O
fake	O
credit	O
card	O
verification	O
using	O
stolen	O
branding	O
Some	O
of	O
the	O
campaigns	O
appear	O
to	O
have	O
a	O
wider	O
reach	O
based	O
on	O
bit.ly	B-Indicator
statistics	O
like	O
this	O
one	O
from	O
October	O
13	O
,	O
2017	O
:	O
Figure	O
16	O
:	O
bit.ly	B-Indicator
statistics	O
for	O
an	O
October	O
13	O
,	O
2017	O
campaign	O
Over	O
several	O
days	O
during	O
the	O
last	O
three	O
months	O
,	O
Proofpoint	B-Organization
researchers	O
observed	O
campaigns	O
using	O
similar	O
techniques	O
targeting	O
the	O
banking	O
customers	O
of	O
Raffeisen	O
and	O
Sparkasse	O
.	O

A	O
review	O
of	O
the	O
bit.ly	B-Indicator
statistics	O
for	O
these	O
campaigns	O
shows	O
that	O
they	O
were	O
at	O
least	O
as	O
effective	O
in	O
driving	O
end-user	O
clicks	O
as	O
the	O
Bank	B-System
Austria	I-System
campaign	O
analyzed	O
above	O
.	O

Conclusion	O
As	O
our	O
computing	O
increasingly	O
crosses	O
multiple	O
screens	O
,	O
we	O
should	O
expect	O
to	O
see	O
threats	O
extending	O
across	O
mobile	O
and	O
desktop	O
environments	O
.	O

Moreover	O
,	O
as	O
we	O
use	O
mobile	O
devices	O
to	O
access	O
the	O
web	O
and	O
phishing	O
templates	O
extend	O
to	O
mobile	O
environments	O
,	O
we	O
should	O
expect	O
to	O
see	O
a	O
greater	O
variety	O
of	O
integrated	O
threats	O
like	O
the	O
scheme	O
we	O
detail	O
here	O
.	O

As	O
on	O
the	O
desktop	O
,	O
mobile	O
users	O
need	O
to	O
be	O
wary	O
of	O
installing	O
applications	O
from	O
outside	O
of	O
legitimate	O
app	O
stores	O
and	O
sources	O
and	O
be	O
on	O
the	O
lookout	O
for	O
bogus	O
banking	O
sites	O
that	O
ask	O
for	O
more	O
information	O
than	O
users	O
would	O
normally	O
provide	O
on	O
legitimate	O
sites	O
.	O

Unusual	O
domains	O
,	O
the	O
use	O
of	O
URL	O
shorteners	O
,	O
and	O
solicitations	O
that	O
do	O
not	O
come	O
from	O
verifiable	O
sources	O
are	O
also	O
red	O
flags	O
for	O
potential	O
phishing	O
and	O
malware	O
.	O

Ginp	B-Malware
-	O
A	O
malware	O
patchwork	O
borrowing	O
from	O
Anubis	B-Malware
November	O
2019	O
Intro	O
ThreatFabric	B-System
analysts	O
have	O
recently	O
investigated	O
an	O
interesting	O
new	O
strain	O
of	O
banking	O
malware	O
.	O

The	O
malware	O
was	O
first	O
spotted	O
by	O
Tatyana	O
Shishkova	O
from	O
Kaspersky	B-Organization
by	O
end	O
October	O
2019	O
,	O
but	O
actually	O
dates	O
back	O
to	O
June	O
2019	O
.	O

It	O
is	O
still	O
under	O
active	O
development	O
,	O
with	O
at	O
least	O
5	O
different	O
versions	O
of	O
the	O
Trojan	O
released	O
within	O
the	O
last	O
5	O
months	O
(	O
June	O
-	O
November	O
2019	O
)	O
.	O

What	O
makes	O
Ginp	B-Malware
stand	O
out	O
is	O
that	O
it	O
was	O
built	O
from	O
scratch	O
being	O
expanded	O
through	O
regular	O
updates	O
,	O
the	O
last	O
of	O
which	O
including	O
code	O
copied	O
from	O
the	O
infamous	O
Anubis	B-Malware
banking	O
Trojan	O
,	O
indicating	O
that	O
its	O
author	O
is	O
cherry-picking	O
the	O
most	O
relevant	O
functionality	O
for	O
its	O
malware	O
.	O

In	O
addition	O
,	O
its	O
original	O
target	O
list	O
is	O
extremely	O
narrow	O
and	O
seems	O
to	O
be	O
focused	O
on	O
Spanish	O
banks	O
.	O

Last	O
but	O
not	O
least	O
,	O
all	O
the	O
overlay	O
screens	O
(	O
injects	O
)	O
for	O
the	O
banks	O
include	O
two	O
steps	O
;	O
first	O
stealing	O
the	O
victim	O
’	O
s	O
login	O
credentials	O
,	O
then	O
their	O
credit	O
card	O
details	O
.	O

Although	O
multi-step	O
overlays	O
are	O
not	O
something	O
new	O
,	O
their	O
usage	O
is	O
generally	O
limited	O
to	O
avoid	O
raising	O
suspicion	O
.	O

Evolution	O
The	O
initial	O
version	O
of	O
the	O
malware	O
dates	O
back	O
to	O
early	O
June	O
2019	O
,	O
masquerading	O
as	O
a	O
“	O
Google	B-System
Play	I-System
Verificator	I-System
”	O
app	O
.	O

At	O
that	O
time	O
,	O
Ginp	B-Malware
was	O
a	O
simple	O
SMS	O
stealer	O
whose	O
purpose	O
was	O
only	O
to	O
send	O
a	O
copy	O
of	O
incoming	O
and	O
outgoing	O
SMS	O
messages	O
to	O
the	O
C2	O
server	O
.	O

A	O
couple	O
of	O
months	O
later	O
,	O
in	O
August	O
2019	O
,	O
a	O
new	O
version	O
was	O
released	O
with	O
additional	O
banking-specific	O
features	O
.	O

This	O
and	O
following	O
versions	O
were	O
masquerading	O
as	O
fake	O
“	O
Adobe	B-System
Flash	I-System
Player	I-System
”	O
apps	O
.	O

The	O
malware	O
was	O
able	O
to	O
perform	O
overlay	O
attacks	O
and	O
become	O
the	O
default	O
SMS	O
app	O
through	O
the	O
abuse	O
of	O
the	O
Accessibility	O
Service	O
.	O

The	O
overlay	O
consisted	O
of	O
a	O
generic	O
credit	O
card	O
grabber	O
targeting	O
social	O
and	O
utility	O
apps	O
,	O
such	O
as	O
Google	B-System
Play	I-System
,	O
Facebook	B-System
,	O
WhatsApp	B-System
,	O
Chrome	B-System
,	O
Skype	B-System
,	O
Instagram	B-System
and	O
Twitter	B-System
.	O

Although	O
early	O
versions	O
had	O
some	O
basic	O
code	O
and	O
string	O
obfuscation	O
,	O
protection	O
of	O
the	O
third	O
version	O
of	O
the	O
malware	O
was	O
enhanced	O
with	O
the	O
use	O
of	O
payload	O
obfuscation	O
.	O

The	O
capabilities	O
remained	O
unchanged	O
,	O
but	O
a	O
new	O
endpoint	O
was	O
added	O
to	O
the	O
Trojan	O
C2	O
allowing	O
it	O
to	O
handle	O
the	O
generic	O
card	O
grabber	O
overlay	O
and	O
specific	O
target	O
overlays	O
(	O
banking	O
apps	O
)	O
separately	O
.	O

In	O
addition	O
,	O
the	O
credit	O
card	O
grabber	O
target	O
list	O
was	O
expanded	O
with	O
Snapchat	B-System
and	O
Viber	B-System
.	O

In	O
the	O
third	O
version	O
spotted	O
in	O
the	O
wild	O
,	O
the	O
author	O
introduced	O
parts	O
of	O
the	O
source	O
code	O
of	O
the	O
infamous	O
Anubis	B-Malware
Trojan	O
(	O
which	O
was	O
leaked	O
earlier	O
in	O
2019	O
)	O
.	O

This	O
change	O
came	O
hand	O
in	O
hand	O
with	O
a	O
new	O
overlay	O
target	O
list	O
,	O
no	O
longer	O
targeting	O
social	O
apps	O
,	O
but	O
focusing	O
on	O
banking	O
instead	O
.	O

A	O
remarkable	O
fact	O
is	O
that	O
all	O
the	O
targeted	O
apps	O
relate	O
to	O
Spanish	O
banks	O
,	O
including	O
targets	O
never	O
seen	O
before	O
in	O
any	O
other	O
Android	B-System
banking	O
Trojan	O
.	O

The	O
24	O
target	O
apps	O
belong	O
to	O
7	O
different	O
Spanish	O
banks	O
:	O
Caixa	B-System
bank	I-System
,	O
Bankinter	B-System
,	O
Bankia	B-System
,	O
BBVA	B-System
,	O
EVO	B-System
Banco	I-System
,	O
Kutxabank	B-System
and	O
Santander	B-System
.	O

The	O
specific	O
apps	O
can	O
be	O
found	O
in	O
the	O
target	O
list	O
in	O
the	O
appendix	O
.	O

The	O
most	O
recent	O
version	O
of	O
Ginp	B-Malware
(	O
at	O
the	O
time	O
of	O
writing	O
)	O
was	O
detected	O
at	O
the	O
end	O
of	O
November	O
2019	O
.	O

This	O
version	O
has	O
some	O
small	O
modifications	O
which	O
seems	O
to	O
be	O
unused	O
,	O
as	O
the	O
malware	O
behaviour	O
is	O
the	O
same	O
as	O
the	O
previous	O
version	O
.	O

The	O
author	O
has	O
introduced	O
the	O
capability	O
to	O
grant	O
the	O
app	O
the	O
device	O
admin	O
permission	O
.	O

Additionally	O
new	O
endpoint	O
was	O
added	O
that	O
seems	O
related	O
to	O
downloading	O
a	O
module	O
for	O
the	O
malware	O
,	O
probably	O
with	O
new	O
features	O
or	O
configuration	O
.	O

How	O
it	O
works	O
When	O
the	O
malware	O
is	O
first	O
started	O
on	O
the	O
device	O
it	O
will	O
begin	O
by	O
removing	O
its	O
icon	O
from	O
the	O
app	O
drawer	O
,	O
hiding	O
from	O
the	O
end	O
user	O
.	O

In	O
the	O
second	O
step	O
it	O
asks	O
the	O
victim	O
for	O
the	O
Accessibility	O
Service	O
privilege	O
as	O
visible	O
in	O
following	O
screenshot	O
:	O
Ginp	B-Malware
Accessibility	O
request	O
Once	O
the	O
user	O
grants	O
the	O
requested	O
Accessibility	O
Service	O
privilege	O
,	O
Ginp	B-Malware
starts	O
by	O
granting	O
itself	O
additional	O
permissions	O
,	O
such	O
as	O
(	O
dynamic	O
)	O
permissions	O
required	O
in	O
order	O
to	O
be	O
able	O
to	O
send	O
messages	O
and	O
make	O
calls	O
,	O
without	O
requiring	O
any	O
further	O
action	O
from	O
the	O
victim	O
.	O

When	O
done	O
,	O
the	O
bot	O
is	O
functional	O
and	O
ready	O
to	O
receive	O
commands	O
and	O
perform	O
overlay	O
attacks	O
.	O

The	O
commands	O
supported	O
by	O
the	O
most	O
recent	O
version	O
of	O
the	O
bot	O
are	O
listed	O
below	O
.	O

As	O
can	O
be	O
observed	O
,	O
the	O
possibilities	O
offered	O
by	O
the	O
bot	O
are	O
pretty	O
common	O
.	O

Command	O
Description	O
SEND_SMS	O
Send	O
an	O
SMS	O
from	O
the	O
bot	O
to	O
a	O
specific	O
number	O
NEW_URL	O
Update	O
the	O
C2	O
URL	O
KILL	O
Disable	O
the	O
bot	O
PING_DELAY	O
Update	O
interval	O
between	O
each	O
ping	O
request	O
CLEAN_IGNORE_PKG	O
Empty	O
list	O
of	O
overlayed	O
apps	O
WRITE_INJECTS	O
Update	O
target	O
list	O
READ_INJECTS	O
Get	O
current	O
target	O
list	O
START_ADMIN	O
Request	O
Device	O
Admin	O
privileges	O
ALL_SMS	O
Get	O
all	O
SMS	O
messages	O
DISABLE_ACCESSIBILITY	O
Stop	O
preventing	O
user	O
from	O
disabling	O
the	O
accessibility	O
service	O
ENABLE_ACCESSIBILITY	O
Prevent	O
user	O
from	O
disabling	O

the	O
accessibility	O
service	O
ENABLE_HIDDEN_SMS	O
Set	O
malware	O
as	O
default	O
SMS	O
app	O
DISABLE_HIDDEN_SMS	O
Remove	O
malware	O
as	O
default	O
SMS	O
app	O
ENABLE_EXTENDED_INJECT	O
Enable	O
overlay	O
attacks	O
DISABLE_EXTENDED_INJECT	O
Disable	O
overlay	O
attacks	O
ENABLE_CC_GRABBER	O
Enable	O
the	O
Google	B-System
Play	I-System
overlay	O
DISABLE_CC_GRABBER	O
Disable	O
the	O
Google	B-System
Play	I-System
overlay	O
START_DEBUG	O
Enable	O
debugging	O
GET_LOGCAT	O
Get	O
logs	O
from	O
the	O
device	O
STOP_DEBUG	O
Disable	O
debugging	O
GET_APPS	O

Get	O
installed	O
applications	O
GET_CONTACTS	O
Get	O
contacts	O
SEND_BULK_SMS	O
Send	O
SMS	O
to	O
multiple	O
numbers	O
UPDATE_APK	O
Not	O
implemented	O
INJECT_PACKAGE	O
Add	O
new	O
overlay	O
target	O
CALL_FORWARD	O
Enable/disable	O
call	O
forwarding	O
START_PERMISSIONS	O
Starts	O
request	O
for	O
additional	O
permissions	O
(	O
Accessibility	O
privileges	O
,	O
battery	O
optimizations	O
bypass	O
,	O
dynamic	O
permissions	O
)	O
Features	O
The	O
most	O
recent	O
version	O
of	O
Ginp	O
has	O
the	O
same	O
capabilities	O
as	O
most	O
other	O
Android	B-System
banking	O
Trojans	O
,	O
such	O
as	O
the	O
use	O
of	O
overlay	O
attacks	O
,	O
SMS	O
control	O
and	O
contact	O

list	O
harvesting	O
.	O

Overall	O
,	O
it	O
has	O
a	O
fairly	O
common	O
feature	O
list	O
,	O
but	O
it	O
is	O
expected	O
to	O
expand	O
in	O
future	O
updates	O
.	O

Since	O
Ginp	O
is	O
already	O
using	O
some	O
code	O
from	O
the	O
Anubis	B-Malware
Trojan	O
,	O
it	O
is	O
quite	O
likely	O
that	O
other	O
,	O
more	O
advanced	O
features	O
from	O
Anubis	B-System
or	O
other	O
malware	O
,	O
such	O
as	O
a	O
back-connect	O
proxy	O
,	O
screen-streaming	O
and	O
RAT	O
will	O
also	O
be	O
added	O
in	O
the	O
future	O
.	O

Ginp	B-Malware
embeds	O
the	O
following	O
set	O
of	O
features	O
,	O
allowing	O
it	O
to	O
remain	O
under	O
the	O
radar	O
and	O
successfully	O
perform	O
attacks	O
:	O
Overlaying	O
:	O
Dynamic	O
(	O
local	O
overlays	O
obtained	O
from	O
the	O
C2	O
)	O
SMS	O
harvesting	O
:	O
SMS	O
listing	O
SMS	O
harvesting	O
:	O
SMS	O
forwarding	O
Contact	O
list	O
collection	O
Application	O
listing	O
Overlaying	O
:	O
Targets	O
list	O
update	O
SMS	O
:	O
Sending	O
Calls	O
:	O
Call	O
forwarding	O
C2	O
Resilience	O
:	O
Auxiliary	O
C2	O
list	O
Self-protection	O
:	O
Hiding	O
the	O
App	O
icon	O
Self-protection	O
:	O
Preventing	O
removal	O
Self-protection	O
:	O
Emulation-detection	O
Update	O

10/03/2020	O
At	O
the	O
end	O
of	O
February	O
the	O
actors	O
behind	O
Ginp	B-Malware
added	O
screen	O
capture	O
capabilities	O
to	O
their	O
Trojan	O
.	O

Like	O
previously	O
added	O
functionality	O
,	O
the	O
code	O
is	O
borrowed	O
from	O
the	O
leaked	O
Anubis	B-Malware
Trojan	O
source	O
code	O
.	O

It	O
enables	O
the	O
bot	O
to	O
stream	O
screenshots	O
and	O
send	O
them	O
to	O
the	O
C2	O
so	O
that	O
actors	O
can	O
see	O
what	O
is	O
happening	O
on	O
the	O
screen	O
of	O
the	O
infected	O
device	O
.	O

Overlay	O
attack	O
Ginp	O
uses	O
the	O
Accessibility	O
Service	O
to	O
check	O
which	O
application	O
runs	O
is	O
the	O
foreground	O
.	O

If	O
the	O
package	O
name	O
of	O
the	O
foreground	O
app	O
is	O
included	O
in	O
the	O
target	O
list	O
,	O
an	O
overlay	O
is	O
shown	O
.	O

The	O
WebView-based	O
overlay	O
is	O
loading	O
an	O
HTML	O
page	O
provided	O
by	O
the	O
C2	O
in	O
response	O
to	O
the	O
package	O
name	O
provided	O
by	O
the	O
bot	O
.	O

Something	O
that	O
makes	O
Ginp	B-Malware
special	O
is	O
that	O
all	O
of	O
its	O
overlay	O
screens	O
for	O
banking	O
apps	O
are	O
consist	O
of	O
multiple	O
steps	O
,	O
first	O
stealing	O
the	O
victim	O
’	O
s	O
login	O
credentials	O
,	O
then	O
stealing	O
the	O
credit	O
card	O
details	O
(	O
to	O
“	O
validate	O
”	O
the	O
user	O
identity	O
)	O
,	O
as	O
shown	O
in	O
the	O
screenshots	O
hereafter	O
:	O
The	O
following	O
code	O
snippet	O
shows	O
that	O
after	O
the	O
second	O
overlay	O
is	O
filled-in	O
and	O
validated	O
,	O
it	O
disappears	O
and	O
the	O
targeted	O
application	O
is	O
added	O
to	O
the	O
list	O
of	O
packages	O
names	O
to	O
be	O
ignored	O
for	O
future	O
overlays	O
attacks	O
.	O

Targets	O
The	O
initial	O
version	O
of	O
Ginp	B-Malware
had	O
a	O
generic	O
credit	O
card	O
grabber	O
overlay	O
screen	O
used	O
for	O
all	O
targeted	O
applications	O
.	O

Still	O
included	O
in	O
the	O
last	O
versions	O
,	O
this	O
screen	O
is	O
only	O
used	O
to	O
overlay	O
the	O
official	O
Google	B-System
Play	I-System
Store	I-System
app	O
.	O

More	O
apps	O
could	O
be	O
added	O
to	O
the	O
grabber	O
target	O
list	O
in	O
the	O
future	O
,	O
such	O
as	O
the	O
ones	O
that	O
were	O
targeted	O
in	O
older	O
versions	O
:	O
Facebook	B-System
WhatsApp	B-System
Skype	B-System
Twitter	B-System
Chrome	B-System
Instagram	B-System
Snapchat	B-System
Viber	B-System
The	O
following	O
screenshot	O
shows	O
the	O
generic	O
card	O
grabber	O
overlay	O
screen	O
:	O
Ginp	B-Malware
generic	O
grabber	O
The	O
current	O
active	O
target	O
list	O
is	O
available	O
in	O
the	O
appendix	O
,	O
containing	O
a	O
total	O
of	O
24	O
unique	O
targets	O
.	O

The	O
following	O
screenshots	O
show	O
what	O
type	O
of	O
information	O
is	O
collected	O
in	O
both	O
steps	O
of	O
the	O
overlay	O
attack	O
:	O
Ginp	B-Malware
overlaysGinp	O
overlaysGinp	O
overlaysGinp	O
overlays	O
Based	O
on	O
Anubis	B-Malware
Once	O
the	O
Anubis	B-Malware
bot	O
code	O
got	O
leaked	O
,	O
it	O
was	O
just	O
a	O
matter	O
of	O
time	O
before	O
new	O
banking	O
Trojans	O
based	O
on	O
Anubis	B-Malware
would	O
surface	O
.	O

When	O
analyzing	O
the	O
Ginp	B-Malware
’	O
s	O
recent	O
samples	O
,	O
ThreatFabric	B-System
analysts	O
found	O
some	O
similarities	O
with	O
the	O
famous	O
Android	O
banking	O
Trojan	O
.	O

Based	O
on	O
the	O
evolution	O
of	O
Ginp	B-Malware
it	O
is	O
clear	O
that	O
it	O
isn	O
’	O
t	O
based	O
on	O
Anubis	B-Malware
,	O
but	O
rather	O
reuses	O
some	O
of	O
its	O
code	O
.	O

Below	O
are	O
some	O
of	O
the	O
elements	O
showing	O
the	O
relation	O
.	O

The	O
names	O
used	O
for	O
Android	B-System
components	O
are	O
similar	O
:	O
Similarities	O
with	O
AnubisSimilarities	O
with	O
Anubis	B-Malware
When	O
analyzing	O
these	O
components	O
,	O
similarities	O
were	O
found	O
in	O
the	O
code	O
of	O
both	O
malware	O
families	O
:	O
Similarities	O
with	O
Anubis	B-System
Another	O
major	O
change	O
that	O
indicated	O
that	O
the	O
actor	O
copied	O
code	O
from	O
the	O
Anubis	B-Malware
Trojan	O
is	O
the	O
way	O
of	O
handling	O
configuration	O
values	O
.	O

Previous	O
versions	O
were	O
storing	O
config	O
values	O
within	O
the	O
variables	O
of	O
a	O
class	O
,	O
while	O
the	O
latest	O
version	O
is	O
using	O
SharedPreferences	O
with	O
some	O
of	O
the	O
keys	O
being	O
identical	O
to	O
those	O
used	O
by	O
Anubis	B-System
:	O
isAccessibility	O
time_work	O
time_start_permission	O
url_inj	O
Conclusion	O
Ginp	B-Malware
is	O
a	O
simple	O
but	O
rather	O
efficient	O
banking	O
Trojan	O
providing	O
the	O
basic	O
functionality	O
to	O
be	O
able	O
to	O
trick	O
victims	O
into	O
delivering	O
personal	O
information	O
.	O

In	O
a	O
5-month	O
timespan	O
,	O
actor	O
managed	O
to	O
create	O
a	O
Trojan	O
from	O
scratch	O
which	O
will	O
presumably	O
continue	O
evolving	O
offering	O
new	O
features	O
such	O
as	O
keylogging	O
,	O
back-connect	O
proxy	O
or	O
RAT	O
capabilities	O
.	O

Ginp	B-Malware
’	O
s	O
unusual	O
target	O
selection	O
is	O
not	O
just	O
about	O
its	O
focus	O
on	O
Spanish	O
banks	O
but	O
also	O
the	O
wide	O
selection	O
of	O
targeted	O
apps	O
per	O
bank	O
.	O

The	O
fact	O
that	O
the	O
overlay	O
screens	O
are	O
almost	O
identical	O
to	O
the	O
legitimate	O
banking	O
apps	O
suggests	O
that	O
the	O
actors	O
might	O
be	O
very	O
familiar	O
with	O
the	O
Spanish	O
banking	O
applications	O
and	O
might	O
even	O
be	O
accustomed	O
to	O
the	O
language	O
.	O

Although	O
the	O
current	O
target	O
list	O
is	O
limited	O
to	O
Spanish	O
apps	O
,	O
it	O
seems	O
that	O
the	O
actor	O
is	O
taking	O
into	O
account	O
that	O
the	O
bot	O
should	O
also	O
be	O
able	O
to	O
target	O
other	O
countries	O
,	O
seeing	O
that	O
the	O
path	O
used	O
in	O
the	O
inject	O
requests	O
contains	O
the	O
country	O
code	O
of	O
the	O
targeted	O
institution	O
.	O

This	O
could	O
indicate	O
that	O
actor	O
already	O
has	O
plans	O
in	O
expanding	O
the	O
targets	O
to	O
applications	O
from	O
different	O
countries	O
and	O
regions	O
.	O

Appendix	O
Samples	O
Some	O
of	O
the	O
latest	O
Ginp	B-Malware
samples	O
found	O
in	O
the	O
wild	O
:	O
App	O
name	O
Package	O
name	O
SHA-256	O
hash	O
Google	B-System
Play	I-System
Verificator	I-System
sing.guide.false	B-Indicator
0ee075219a2dfde018f17561467272633821d19420c08cba14322cc3b93bb5d5	B-Indicator
Google	B-System
Play	I-System
Verificator	I-System
park.rather.dance	B-System
087a3beea46f3d45649b7506073ef51c784036629ca78601a4593759b253d1b7	B-Indicator
Adobe	B-System
Flash	I-System
Player	I-System
ethics.unknown.during	B-Indicator

5ac6901b232c629bc246227b783867a0122f62f9e087ceb86d83d991e92dba2f	B-Indicator
Adobe	B-System
Flash	I-System
Player	I-System
solution.rail.forward	B-Indicator
7eb239cc86e80e6e1866e2b3a132b5af94a13d0d24f92068a6d2e66cfe5c2cea	B-Indicator
Adobe	B-System
Flash	I-System
Player	I-System
com.pubhny.hekzhgjty	B-Indicator
14a1b1dce69b742f7e258805594f07e0c5148b6963c12a8429d6e15ace3a503c	B-Indicator

Adobe	B-System
Flash	I-System
Player	I-System
sentence.fancy.humble	B-Indicator
78557094dbabecdc17fb0edb4e3a94bae184e97b1b92801e4f8eb0f0626d6212	B-Indicator
Target	O
list	O
The	O
current	O
list	O
of	O
apps	O
observed	O
to	O
be	O
targeted	O
by	O
Ginp	B-Malware
contains	O
a	O
total	O
of	O
24	O
unique	O
applications	O
as	O
seen	O
below	O
.	O

This	O
list	O
is	O
expected	O
to	O
grow	O
in	O
the	O
future	O
.	O