WEBVTT

00:00:00.240 --> 00:00:00.960
‑Hey humans.

00:00:00.970 --> 00:00:02.730
I am here today with Dr.

00:00:02.730 --> 00:00:06.590
Josh Stroschein, and we're going to talk about the Dridex Banking Trojan,

00:00:06.590 --> 00:00:11.490
ways that we can analyze its behavior to understand how to detect it,

00:00:11.490 --> 00:00:12.050
prevent it,

00:00:12.050 --> 00:00:16.350
and respond in case we get infected with that or we find it in our enterprise.

00:00:17.040 --> 00:00:19.510
So, thanks, Josh, for being here today, I really appreciate it.

00:00:19.510 --> 00:00:20.760
I'm excited to jump into this.

00:00:21.740 --> 00:00:22.540
‑Yeah, thanks, Aaron.

00:00:22.550 --> 00:00:26.640
I am also excited to be here and for all of you to be joining us as well.

00:00:26.650 --> 00:00:29.710
To get started then, I've just got a few slides,

00:00:29.710 --> 00:00:32.009
provide an overview, and then a brief introduction here,

00:00:32.009 --> 00:00:34.220
and then we'll get into a lot of demonstration and hands on.

00:00:34.220 --> 00:00:39.160
So, we'll start with an introduction to Dridex, what it is, and what does it do.

00:00:39.540 --> 00:00:42.290
We'll then move into three primary categories,

00:00:42.300 --> 00:00:46.130
preventing Dridex, we'll look at some tips and techniques for you,

00:00:46.140 --> 00:00:49.020
helping you to prevent an infection, we'll talk about detection,

00:00:49.020 --> 00:00:51.410
how can you identify key behaviors,

00:00:51.410 --> 00:00:54.720
and finding those ever‑important indicators of compromise.

00:00:54.730 --> 00:00:56.290
We'll look at response,

00:00:56.300 --> 00:00:59.660
discuss some tasks that you should consider if in the

00:00:59.660 --> 00:01:01.750
unfortunate case an infection occurs.

00:01:02.640 --> 00:01:05.870
Right, so with that, I think we're ready to get into our introduction.

00:01:05.880 --> 00:01:08.960
So a real brief background on Dridex.

00:01:08.970 --> 00:01:11.740
Dridex has been around for quite some time.

00:01:11.740 --> 00:01:13.690
It was first seen in early 2014,

00:01:13.690 --> 00:01:16.020
and if you really get into the history of Dridex,

00:01:16.030 --> 00:01:19.800
you'll see that it has evolved from some previous malware families.

00:01:19.800 --> 00:01:24.000
So it has some genealogy and some code patterns that have stayed really

00:01:24.000 --> 00:01:28.460
persistent as it's grown into the modern threat that it is today.

00:01:29.240 --> 00:01:34.390
Dridex has a very modular design, and it has been consistently kept up to date,

00:01:34.390 --> 00:01:37.010
which is one of the reasons why we're discussing it here.

00:01:37.020 --> 00:01:40.540
The modular design then is something to keep in mind in that

00:01:40.540 --> 00:01:44.540
while it is primarily a banking Trojan, because it is modular,

00:01:44.540 --> 00:01:46.720
it does allow it to bring in additional capabilities,

00:01:46.720 --> 00:01:51.260
and so we really need to consider this or view Dridex as an all‑purpose threat.

00:01:51.440 --> 00:01:54.370
It is delivered primarily through malicious Office

00:01:54.370 --> 00:01:57.460
documents using VBA and Excel 4‑style macros.

00:01:57.460 --> 00:02:00.190
We'll get into those in a bit more detail here in just a bit.

00:02:00.200 --> 00:02:04.270
And as I said, it's primarily targeting banking information.

00:02:04.280 --> 00:02:06.350
It's able to perform web injects,

00:02:06.350 --> 00:02:08.940
so stealing credentials directly out of the browser,

00:02:08.949 --> 00:02:10.810
performing key logging,

00:02:10.810 --> 00:02:13.510
as well as even taking screen captures to get that information,

00:02:13.510 --> 00:02:16.620
but it doesn't limit itself to just that purpose.

00:02:16.620 --> 00:02:20.450
With that modular nature, it allows it to expand its capabilities.

00:02:21.240 --> 00:02:23.760
‑I have a few questions based on this and some other

00:02:23.760 --> 00:02:26.660
malware that we've talked about in the past.

00:02:26.670 --> 00:02:30.260
So the first thing is we talk about banking Trojans,

00:02:30.260 --> 00:02:33.660
and I want to be clear that when we talk about what industry vertical that is,

00:02:34.040 --> 00:02:36.260
we're talking about the customers for the bank,

00:02:36.740 --> 00:02:41.370
right, this isn't malware specifically designed to attack Bank of America,

00:02:41.370 --> 00:02:43.510
or JPMorgan, and Chase at their enterprise,

00:02:43.510 --> 00:02:45.840
this is to attack the customers at their house,

00:02:45.840 --> 00:02:46.660
is that accurate?

00:02:47.240 --> 00:02:49.740
‑Yeah, yeah, that's that is absolutely correct.

00:02:49.750 --> 00:02:53.470
The primary focus is getting in front of the everyday

00:02:53.470 --> 00:02:55.710
user to steal their banking credentials,

00:02:55.710 --> 00:03:01.460
to then steal money or whatever nefarious deeds they want

00:03:01.460 --> 00:03:04.520
to perpetuate with those credentials,

00:03:04.520 --> 00:03:07.470
and then also to make those systems part of this

00:03:07.470 --> 00:03:09.700
larger botnet that then they become,

00:03:09.700 --> 00:03:13.820
they can be utilized for spreading additional malware and other,

00:03:13.820 --> 00:03:15.160
again other malicious activity.

00:03:15.540 --> 00:03:17.650
‑Oh, I see, yeah, and then much like some other malware,

00:03:17.650 --> 00:03:22.280
we've seen the trade access to other organizations or sell it off,

00:03:22.290 --> 00:03:23.140
that kind of deal.

00:03:23.140 --> 00:03:24.360
‑Yeah, they certainly can.

00:03:25.840 --> 00:03:30.820
So it's important to think about where it's most relevant or where the

00:03:30.820 --> 00:03:34.130
greatest opportunity is to prevent an infection from Dridex,

00:03:34.140 --> 00:03:37.650
and with Dridex, as with many modern threats,

00:03:37.660 --> 00:03:40.650
stopping the infection from the very beginning is

00:03:40.660 --> 00:03:43.340
really the most important step.

00:03:43.350 --> 00:03:45.210
As I previously noted,

00:03:45.220 --> 00:03:50.420
Dridex is delivered primarily through Office documents using Word and Excel,

00:03:50.420 --> 00:03:54.010
while it hasn't, I haven't observed any PowerPoint usage,

00:03:54.010 --> 00:03:56.170
PowerPoint has the same macro capability,

00:03:56.170 --> 00:03:57.350
so it is feasible.

00:03:57.350 --> 00:03:59.350
‑I was going to say, these macros,

00:03:59.350 --> 00:04:01.350
they're a little bit different because it's not

00:04:01.350 --> 00:04:03.310
just necessarily just like a .vbs.

00:04:03.310 --> 00:04:08.330
like the 4.0 macros are like a language that's kind of built into

00:04:08.340 --> 00:04:10.580
the Office documents or the Office capability,

00:04:10.580 --> 00:04:11.360
is that accurate?

00:04:12.040 --> 00:04:12.930
‑Yeah, yeah.

00:04:12.930 --> 00:04:18.060
The difference with, and why we differentiate between VBA and Excel 4,

00:04:18.060 --> 00:04:21.459
Excel 4 macros of course are just limited to Excel.

00:04:21.470 --> 00:04:26.080
They've been deprecated, and like many things that we see with Microsoft,

00:04:26.090 --> 00:04:28.610
even though they've been deprecated for the newer style,

00:04:28.610 --> 00:04:31.260
the VBA style, Visual Basic for Applications,

00:04:31.540 --> 00:04:33.160
the OfficeSuite still supports them.

00:04:33.440 --> 00:04:37.010
And so what's happened is that as they've kind of fallen

00:04:37.010 --> 00:04:38.980
off the radar over the last few years,

00:04:38.980 --> 00:04:42.510
Excel 4 macros I think were created back in the '90s,

00:04:42.510 --> 00:04:44.160
they've been forgotten,

00:04:44.160 --> 00:04:46.290
and a lot of our detection and our tooling has been

00:04:46.290 --> 00:04:48.620
then built around the VBA style.

00:04:48.620 --> 00:04:51.170
Because then they've been overlooked for so long,

00:04:51.180 --> 00:04:54.460
when they were sort of reintroduced a couple of years ago,

00:04:54.640 --> 00:04:56.050
they were stealthy,

00:04:56.060 --> 00:04:58.910
and they were getting and kind of flying under the radar

00:04:58.910 --> 00:05:00.320
of a lot of our security products,

00:05:00.320 --> 00:05:04.260
and so they were quite effective of getting through the perimeter,

00:05:04.260 --> 00:05:06.910
getting on the host, and then getting in front of the user,

00:05:06.910 --> 00:05:09.180
so that they would hopefully enable the content.

00:05:09.180 --> 00:05:11.390
Once they're in front of the user, it's very similar,

00:05:11.390 --> 00:05:15.650
the enable content type of a ribbon, as as you see here,

00:05:15.840 --> 00:05:18.070
they have to rely on the social engineering.

00:05:18.080 --> 00:05:21.100
There really isn't a lot of exploit use.

00:05:21.100 --> 00:05:26.180
A few years ago, 2017, there was the Equation Editor that was,

00:05:26.190 --> 00:05:29.280
and it's still used to this day, but it was quite heavily used at the time.

00:05:29.290 --> 00:05:33.050
And the power of that was that you didn't have to get

00:05:33.050 --> 00:05:34.530
the user to interact with the doc,

00:05:34.530 --> 00:05:37.870
you just were able to exploit it once they opened it on the system.

00:05:37.870 --> 00:05:40.950
With the VBA and even the Excel 4,

00:05:40.960 --> 00:05:45.670
most organizations are set up to block the initial execution of macros,

00:05:45.680 --> 00:05:49.570
but they get the option, they still give the user the option to enable content.

00:05:49.580 --> 00:05:53.490
And so they really rely heavily, and this isn't unique to just Dridex,

00:05:53.490 --> 00:05:56.170
many threat actors that are distributing through Maldoc,

00:05:56.170 --> 00:05:58.850
they rely very heavily on the social engineering.

00:05:59.740 --> 00:06:02.040
‑Yeah, I think there's two awesome points,

00:06:02.040 --> 00:06:04.230
as one, it's still users clicking buttons,

00:06:04.240 --> 00:06:08.650
right, and then two, it's basically,

00:06:08.650 --> 00:06:13.700
it's so old that we don't detect on it anymore,

00:06:13.710 --> 00:06:15.690
like Excel 4.0 because it's been deprecating,

00:06:15.690 --> 00:06:19.380
and so I would like to make a point on this one about just

00:06:19.380 --> 00:06:23.270
because it's old doesn't mean it's more secure because there's

00:06:23.270 --> 00:06:26.300
some groups that like to say like, oh, it's so old and secure,

00:06:26.300 --> 00:06:29.030
like a mainframe, right, and that's not necessarily true.

00:06:29.030 --> 00:06:31.180
It's so old it may not be in our signatures at that point.

00:06:31.190 --> 00:06:32.660
So, awesome.

00:06:33.040 --> 00:06:33.690
‑Yeah, yeah,

00:06:33.690 --> 00:06:36.670
and while we don't have time to get into all of the

00:06:36.670 --> 00:06:39.170
intricacies of how these documents are written,

00:06:39.180 --> 00:06:43.270
something that I've observed with the Excel 4 macros is

00:06:43.280 --> 00:06:46.120
there is a lot more anti‑analysis in them.

00:06:46.120 --> 00:06:50.490
The code is written in a way to test for certain environmental variables,

00:06:50.490 --> 00:06:52.900
is it a windows host, does the host have a mouse,

00:06:52.900 --> 00:06:54.860
doesn't it have the ability to play sound,

00:06:54.860 --> 00:06:58.590
and that also then contributes to sort of the stealthy nature

00:06:58.600 --> 00:07:00.400
from the threat researcher perspective,

00:07:00.400 --> 00:07:03.060
in that we throw these things in mass in a sandbox,

00:07:03.060 --> 00:07:04.090
they look like they're benign,

00:07:04.090 --> 00:07:06.160
and then they don't get pushed out into the threat feeds

00:07:06.160 --> 00:07:07.750
that a lot of organizations are consuming.

00:07:08.340 --> 00:07:10.760
So here you can see a really good example of that

00:07:10.760 --> 00:07:12.260
social engineering I mentioned.

00:07:12.270 --> 00:07:16.020
Oftentimes, there's going to be brand abuse.

00:07:16.030 --> 00:07:18.210
Here you can see two brands that are being abused.

00:07:18.210 --> 00:07:19.840
Walt Disney World is the primary,

00:07:19.850 --> 00:07:24.150
and then there is the Office theme in the main image in the center.

00:07:24.540 --> 00:07:25.950
Beyond that,

00:07:25.950 --> 00:07:30.360
there's typically some enticement as to why the user needs to enable content.

00:07:30.370 --> 00:07:32.300
In this case, it's because you've got all these,

00:07:32.310 --> 00:07:36.710
this privacy feature that is blocking out certain critical content,

00:07:36.710 --> 00:07:39.180
and this, what looks to be an invoice, a guest invoice.

00:07:39.180 --> 00:07:43.230
So again, the real weak point here in the delivery is just getting,

00:07:43.240 --> 00:07:46.090
once it's passed any sort of email filtering that the

00:07:46.090 --> 00:07:50.370
organization is deploying or utilizing, it's really up to the user,

00:07:50.380 --> 00:07:52.700
or how the environment,

00:07:52.700 --> 00:07:56.400
how the organization has configured the environment to help prevent what

00:07:56.400 --> 00:07:59.720
happens if this malicious Office document gets on the host.

00:07:59.720 --> 00:08:02.550
Once, and if it, so if it gets to the user,

00:08:02.560 --> 00:08:03.420
they enable content,

00:08:03.420 --> 00:08:07.360
it's very typical for these to move into living off the land style techniques,

00:08:07.370 --> 00:08:11.180
that is using built‑in capabilities in the operating system.

00:08:11.180 --> 00:08:14.070
PowerShell has been prevalent now for quite some time,

00:08:14.070 --> 00:08:16.200
rundll32, which we'll talk about today,

00:08:16.200 --> 00:08:20.850
as well as Regsvr32 binaries that they can use on the system

00:08:20.940 --> 00:08:24.120
to help avoid creating additional noise,

00:08:24.130 --> 00:08:26.260
trying to limit what they're writing to the file system,

00:08:26.260 --> 00:08:29.850
trying to limit the tools that they're bringing in.

00:08:29.860 --> 00:08:32.210
So we'll see that in action today.

00:08:32.220 --> 00:08:35.270
There are quite a few host and network‑based indicators,

00:08:35.270 --> 00:08:38.690
and what we're going to explore is how you can identify them to help

00:08:38.690 --> 00:08:41.260
with the identification and with the prevention.

00:08:41.640 --> 00:08:46.440
And then finally, if you're following along with your MITRE ATT&CK bingo sheet,

00:08:46.450 --> 00:08:50.610
this falls under MITRE ATT&CK T1566.001.

00:08:50.610 --> 00:08:55.010
And how we're going to approach the discussion to their conversation is

00:08:55.010 --> 00:08:58.320
really just kind of move through MITRE from left to right,

00:08:58.320 --> 00:09:03.260
from delivery to infection, and some of the post infection activities.

00:09:04.540 --> 00:09:05.190
‑Perfect.

00:09:05.200 --> 00:09:05.590
Yeah.

00:09:05.600 --> 00:09:06.770
Looking forward to it.

00:09:06.780 --> 00:09:09.260
We're saying phishing is really the 100,

00:09:09.260 --> 00:09:12.160
are we seeing anything other than phishing from this malware,

00:09:13.140 --> 00:09:14.550
would be the main question I have.

00:09:15.240 --> 00:09:17.090
‑This is primarily phishing.

00:09:17.090 --> 00:09:18.670
I have not seen much else.

00:09:18.670 --> 00:09:21.410
I can't say 100% that they don't have other techniques,

00:09:21.410 --> 00:09:26.120
but by and large, the vacuum that has been left behind by EMOTET,

00:09:26.120 --> 00:09:29.340
there's a question as to who's going to fill that,

00:09:29.350 --> 00:09:30.600
fill that vacuum.

00:09:30.610 --> 00:09:34.060
Is it going to be actors like Dridex or TrickBot?

00:09:34.070 --> 00:09:35.990
And if so,

00:09:35.990 --> 00:09:40.110
I think that those actors will continue to work in

00:09:40.110 --> 00:09:43.840
the sort of at‑scale techniques, and certainly phishing,

00:09:43.840 --> 00:09:47.270
and phishing campaigns is an effective at‑scale technique

00:09:47.270 --> 00:09:50.260
for getting a large number of victims.

00:09:50.640 --> 00:09:52.090
‑I was going to say that's a really good discussion

00:09:52.090 --> 00:09:55.370
about the vacuum that's left behind, whether it's EMOTET going down,

00:09:55.370 --> 00:09:58.680
or in the future, TrickBot is now taken down,

00:09:58.680 --> 00:09:59.540
and who's feeling that.

00:09:59.540 --> 00:10:03.360
It's not from a perspective of users are really sad because

00:10:03.360 --> 00:10:04.710
they're not getting phished as much anymore,

00:10:04.710 --> 00:10:08.430
that's certainly not it, it's more from an economic standpoint.

00:10:08.430 --> 00:10:11.840
Within black markets, there is a desire to buy access to things,

00:10:11.840 --> 00:10:15.520
there's a desire to buy valid bank emails,

00:10:15.520 --> 00:10:18.620
there's a desire to conduct that criminal activity,

00:10:18.620 --> 00:10:22.060
and that exerts economic pressure that creates a vacuum

00:10:22.440 --> 00:10:25.720
where criminals will then fill that by being the new

00:10:25.720 --> 00:10:27.330
TrickBot or in this case Dridex.

00:10:27.340 --> 00:10:29.460
So, I think that's an excellent conversation.

00:10:30.040 --> 00:10:33.650
‑Yeah, yeah, and with the phishing and the phishing campaigns,

00:10:33.650 --> 00:10:34.840
that that scales well.

00:10:34.850 --> 00:10:38.800
If you look at groups that may be attack organizations,

00:10:38.800 --> 00:10:41.540
they go after infrastructure, they look for exposed RDP,

00:10:41.540 --> 00:10:45.220
that's been an issue with the way that things have changed this last year.

00:10:45.230 --> 00:10:47.080
Well, you can automate that.

00:10:47.080 --> 00:10:49.090
I think there's also an element of that that's a bit more

00:10:49.090 --> 00:10:51.630
manual and maybe even a bit more sophisticated.

00:10:51.630 --> 00:10:57.630
So Maldocs unfortunately are going to be around for a while,

00:10:57.630 --> 00:11:00.390
maybe eventually they'll go the way of the exploit kit,

00:11:00.390 --> 00:11:02.860
but I don't see that happening anytime real soon.

00:11:04.540 --> 00:11:07.480
Moving on to the Trojan itself,

00:11:07.480 --> 00:11:10.670
so the delivery of the Office document is one thing.

00:11:10.680 --> 00:11:14.490
It's one event part of an attack scenario that we

00:11:14.490 --> 00:11:18.960
can focus on and try to prevent, but once that document gets in,

00:11:18.970 --> 00:11:22.220
typically the next thing that happens is some network activity and

00:11:22.220 --> 00:11:24.850
HTTPS requests in order to download the Trojan,

00:11:24.860 --> 00:11:26.100
the loader, and it's modular.

00:11:26.100 --> 00:11:30.120
So, once the loader is there, what we would see is the Dridex Trojan,

00:11:30.120 --> 00:11:32.560
that means other things can eventually follow.

00:11:32.840 --> 00:11:36.320
What's interesting about this is they use a vast

00:11:36.320 --> 00:11:38.040
network of compromised infrastructure.

00:11:38.040 --> 00:11:41.920
That compromise infrastructure is used to host the Trojan,

00:11:41.920 --> 00:11:46.450
and recently that has been a DLL, or a dynamically‑linked library.

00:11:46.460 --> 00:11:50.450
You'll see that it's still a PE file, a Portable Executable.

00:11:50.460 --> 00:11:53.450
There's some activity that we can alert on on the network,

00:11:53.450 --> 00:11:55.400
and we'll explore that here in just a minute.

00:11:55.410 --> 00:11:57.460
In fact, you can see it on the screenshot,

00:11:57.470 --> 00:12:01.360
or in this slide, and then how it's executed,

00:12:01.370 --> 00:12:06.780
executing a DLL change is a little bit of what happens on the host via the

00:12:06.790 --> 00:12:10.360
binaries that then are being utilized in order to execute that.

00:12:10.840 --> 00:12:15.270
The payload typically is requested with a variety of file extensions.

00:12:15.280 --> 00:12:20.400
That is the total red herring, the HTTP request is arbitrary.

00:12:20.410 --> 00:12:23.490
We can create a request for anything as long as the server knows how to

00:12:23.490 --> 00:12:26.740
respond with the content that the requester wants,

00:12:26.750 --> 00:12:31.220
whether it's a .dll, or a .pdf, or a .jpeg,

00:12:31.220 --> 00:12:32.950
or .zip, it really doesn't matter.

00:12:33.040 --> 00:12:36.440
So they use a lot of file extensions, and by and large,

00:12:36.440 --> 00:12:39.340
I assess that to just be a way to try to throw off your analysis.

00:12:39.350 --> 00:12:43.750
The other thing that I think is a bit interesting is that

00:12:43.750 --> 00:12:47.080
all of these DLLs that are on these compromised hosts are

00:12:47.080 --> 00:12:48.680
unobfuscated and unencrypted.

00:12:48.680 --> 00:12:50.140
That means that they're just sitting there.

00:12:50.140 --> 00:12:53.390
While they're in their binary form, they're an executable file,

00:12:53.390 --> 00:12:56.080
there's nothing taken to further obfuscate that.

00:12:56.080 --> 00:12:57.660
That means that when they,

00:12:57.670 --> 00:13:01.470
when the infected system makes the request for this payload,

00:13:01.470 --> 00:13:06.370
then it can be easier to observe when it's coming across the network because oh,

00:13:06.370 --> 00:13:12.850
by the way, they also tend to avoid using HTTPS or a TLS session at this point.

00:13:13.240 --> 00:13:16.360
Okay, so last thing I want to talk about is just a generalized overview.

00:13:16.740 --> 00:13:20.110
Let's lay out the sequence of events that we're going to look at today.

00:13:20.120 --> 00:13:22.220
We start with our malicious Office document,

00:13:22.230 --> 00:13:25.910
Word, Excel are the two predominant types.

00:13:25.920 --> 00:13:28.530
We have macro execution, so social engineering,

00:13:28.530 --> 00:13:30.450
the user, in order to enable that content,

00:13:30.940 --> 00:13:35.640
VBA style or Excel 4, really the end result is the same,

00:13:35.650 --> 00:13:39.770
that is to generate an HTTP request to download the next stage.

00:13:39.780 --> 00:13:44.010
This is typically executed then by rundll32,

00:13:44.010 --> 00:13:47.840
WMIC, it could be PowerShell, this changes a lot,

00:13:47.840 --> 00:13:51.750
and so it is important to really keep kind of up to speed

00:13:51.940 --> 00:13:54.520
with how these documents are evolving.

00:13:54.520 --> 00:13:59.340
Because these are being delivered typically at scale to a number of users,

00:13:59.350 --> 00:14:03.430
we tend to see not significant changes in their structure,

00:14:03.440 --> 00:14:09.500
but regular enough changes that we have to constantly be monitoring how they

00:14:09.500 --> 00:14:12.860
behave in order to understand how best to prevent and detect.

00:14:13.740 --> 00:14:14.900
As I mentioned,

00:14:14.910 --> 00:14:17.400
typically this is where the Trojan will be downloaded

00:14:17.400 --> 00:14:19.620
from a series of compromised hosts.

00:14:19.630 --> 00:14:21.860
That doesn't always have to be the case,

00:14:21.870 --> 00:14:24.250
it's the predominant use case right now.

00:14:24.260 --> 00:14:27.810
It could be that the DLL is actually embedded inside of the document.

00:14:27.810 --> 00:14:31.060
So it could come along inside of that Office doc,

00:14:31.070 --> 00:14:33.450
and then written to disk, and executed there.

00:14:33.540 --> 00:14:34.060
But again,

00:14:34.070 --> 00:14:36.510
the predominant use case right now is that a

00:14:36.510 --> 00:14:38.250
request is made across the internet.

00:14:38.540 --> 00:14:40.480
Now, if it fails to retrieve the Trojan,

00:14:40.810 --> 00:14:43.160
then in a sense the infection has failed.

00:14:43.540 --> 00:14:47.140
You still had a malicious document get in front of a user,

00:14:47.140 --> 00:14:49.970
they still enabled content, so there's still plenty of issues to address,

00:14:49.980 --> 00:14:52.640
but you don't necessarily have Dridex running on that system.

00:14:52.640 --> 00:14:54.830
So again, some indicators,

00:14:54.830 --> 00:14:57.610
some behaviors that we'll be able to look into to help you better

00:14:57.610 --> 00:15:01.360
scope when Dridex has come after your organization.

00:15:01.840 --> 00:15:04.060
And then finally, we move into post infection,

00:15:04.070 --> 00:15:06.180
post infection activity, post infection traffic,

00:15:06.180 --> 00:15:07.310
and this changes.

00:15:07.310 --> 00:15:10.950
Certainly we see, on the network side, we see a change in infrastructure.

00:15:10.960 --> 00:15:13.870
The C2 nodes that it now communicates with are different.

00:15:13.870 --> 00:15:16.590
Typically it's not a quad, typically you see TLS,

00:15:16.600 --> 00:15:19.260
and so everything changes a bit, and we'll explore that as well.

00:15:19.740 --> 00:15:21.260
‑So we have the whole attack chain here.

00:15:21.640 --> 00:15:24.900
When I look from like left‑to‑right on the screen,

00:15:24.910 --> 00:15:27.320
obviously initial access, we have execution,

00:15:27.320 --> 00:15:29.830
we then have like kind of a stage being brought down,

00:15:29.830 --> 00:15:33.090
or a stager being brought down, and then now this is official infection.

00:15:33.100 --> 00:15:36.320
So you're saying, the main thing I wanted to call out is like that,

00:15:36.330 --> 00:15:39.620
that request goes out, if it's blocked in any way on,

00:15:39.620 --> 00:15:41.870
whether it's on the network or on the device from

00:15:41.870 --> 00:15:45.750
executing after it drops back down the DLL,

00:15:45.750 --> 00:15:47.740
you don't, you're not infected.

00:15:47.740 --> 00:15:52.040
There's no like additional logic within that document to tell us,

00:15:52.040 --> 00:15:55.780
okay, let's try again later, or there's nothing dropped into memory to say,

00:15:55.780 --> 00:15:58.980
hey, try this every 5 minutes or anything like that,

00:15:58.980 --> 00:16:01.300
not that you've seen?

00:16:01.300 --> 00:16:02.350
‑Right, not that I've seen,

00:16:02.350 --> 00:16:06.190
and that's not to say it isn't possible because when writing macro code,

00:16:06.190 --> 00:16:08.460
they're powerful, they're capable,

00:16:08.470 --> 00:16:11.690
it is certainly possible that there could be some sort

00:16:11.690 --> 00:16:15.360
of persistence mechanism built there, but generally speaking,

00:16:15.740 --> 00:16:20.430
the macro documents, macro‑enabled documents are really kind of one shot.

00:16:20.440 --> 00:16:23.240
If they get in front of the user, they enable content,

00:16:23.240 --> 00:16:28.540
the macros run, they might try to make a request to one compromised server.

00:16:28.550 --> 00:16:31.250
That's typically what we see with Dridex, that tries one request.

00:16:31.340 --> 00:16:34.110
Some of them might loop through a whole series of domains,

00:16:34.110 --> 00:16:36.270
and it'll try 5, 6, 7 in a row,

00:16:36.270 --> 00:16:39.810
maybe there'll be a little bit of a sleep between the different HTTP requests,

00:16:39.810 --> 00:16:42.250
but ultimately if it fails, then it fails,

00:16:42.250 --> 00:16:42.760
and it's done.

00:16:42.760 --> 00:16:44.680
It exits and it doesn't try again.

00:16:44.680 --> 00:16:48.910
The document itself doesn't try to leave any sort of persistent thing behind.

00:16:48.910 --> 00:16:53.210
If the Trojan's already there embedded in the document,

00:16:53.220 --> 00:16:57.470
and that is able to run, then maybe that actually persists,

00:16:57.470 --> 00:17:01.360
but again, that's not generally what we're seeing with Dridex.

00:17:01.370 --> 00:17:04.079
So yeah, if it fails to make that HTTP request,

00:17:04.079 --> 00:17:08.950
it fails to get the next stage, the actual Dridex loader,

00:17:09.150 --> 00:17:12.349
then you can consider yourself probably lucky at that point.

00:17:13.240 --> 00:17:14.950
‑Yeah, so that gives us a lot of opportunities,

00:17:14.950 --> 00:17:16.339
defenders to look at,

00:17:16.349 --> 00:17:19.910
prevention points in this kill chain that we can start to talk about it,

00:17:19.910 --> 00:17:21.609
and really it's what we're about to talk about next,

00:17:21.609 --> 00:17:22.660
so I'm looking forward to that.
