WEBVTT

00:00:00.140 --> 00:00:01.270
‑Okay, the worst has happened.

00:00:01.280 --> 00:00:04.810
We're getting ready to go home, and we're super excited,

00:00:04.810 --> 00:00:08.220
like a new show just dropped on like Disney+ and we

00:00:08.220 --> 00:00:10.290
want to go watch it with our family, and,

00:00:10.290 --> 00:00:14.400
of course, it's Friday afternoon and we have detected Dridex.

00:00:14.400 --> 00:00:16.059
Because we watched the show in the morning,

00:00:16.059 --> 00:00:18.260
we tried all the cool detections that we just learned about,

00:00:18.740 --> 00:00:20.000
and we actually found stuff,

00:00:20.000 --> 00:00:22.560
and now we have to stay through the weekend and respond.

00:00:22.570 --> 00:00:26.650
What lessons do you have for us about this malware that are going to

00:00:26.660 --> 00:00:29.560
help that individual who is in that situation?

00:00:31.240 --> 00:00:35.000
‑Number of things that are actually quite unique and interesting with

00:00:35.000 --> 00:00:39.130
Dridex and how they deliver not only the Trojan itself,

00:00:39.140 --> 00:00:41.760
but then once that Trojan begins running.

00:00:41.760 --> 00:00:44.970
So we'll look at using some tools like Arkime,

00:00:44.970 --> 00:00:46.590
formerly known as Moloch,

00:00:46.590 --> 00:00:50.230
we'll talk a little bit of maldoc analysis and I'll point you

00:00:50.230 --> 00:00:51.960
to some scripts that are available there,

00:00:51.960 --> 00:00:56.180
and then config extraction and some tools that can help pull out

00:00:56.190 --> 00:00:58.210
additional command and control node information.

00:00:58.210 --> 00:01:01.250
All of that as we look at the response phase.

00:01:02.540 --> 00:01:03.040
‑Excellent.

00:01:03.040 --> 00:01:05.700
And hopefully with these kind of key details,

00:01:05.710 --> 00:01:07.860
we hope that individual will get out of there maybe

00:01:07.860 --> 00:01:09.960
early in the night, or probably not.

00:01:09.970 --> 00:01:12.430
Most likely at least they'll be able to get home before the weekend's

00:01:12.430 --> 00:01:14.880
over if they can leverage some of these capabilities.

00:01:14.880 --> 00:01:16.260
That sounds good.

00:01:16.260 --> 00:01:17.860
What do you do first?

00:01:19.040 --> 00:01:19.250
‑Yeah,

00:01:19.260 --> 00:01:25.290
first let's take a look at ways in which we can discover already known URLs or

00:01:25.290 --> 00:01:28.360
domains that are being utilized to distribute this Trojan.

00:01:28.370 --> 00:01:31.950
‑When we talk about URLs, again, we came into this situation,

00:01:31.950 --> 00:01:32.960
we've detected it,

00:01:32.970 --> 00:01:36.630
and we probably only detected it one or two places now that we're in

00:01:36.630 --> 00:01:39.170
scoping mode. What's the extent of this infection?

00:01:39.170 --> 00:01:41.760
What's the extent of this incident response that

00:01:41.760 --> 00:01:43.420
we're going to have to worry about?

00:01:43.940 --> 00:01:45.850
And so we're going to leverage those URLs that you're

00:01:45.850 --> 00:01:50.460
about to talk about to figure out, hey, let me try to detect it everywhere now.

00:01:50.460 --> 00:01:51.180
‑Right,

00:01:51.180 --> 00:01:56.340
so hopefully an organization has their own internal threat

00:01:56.340 --> 00:01:59.780
intelligence feeds and the data that they can go back,

00:01:59.780 --> 00:02:02.300
network data that they can go back and comb through.

00:02:02.310 --> 00:02:04.650
There's a lot of great resources available online,

00:02:04.650 --> 00:02:08.400
and the URLhaus, part of abuse.ch, is one of those.

00:02:08.400 --> 00:02:12.490
A lot of support from the community, the threat research community,

00:02:12.490 --> 00:02:17.880
to analyze malware families like dridex and get as much

00:02:17.890 --> 00:02:19.620
indicators out there as possible.

00:02:19.620 --> 00:02:21.920
Here you can see an example. We're just going to use

00:02:21.920 --> 00:02:24.630
curl to interact with the URLhaus API.

00:02:24.630 --> 00:02:27.870
We're going to filter by the tag Dridex,

00:02:27.870 --> 00:02:31.970
so what this will provide us with is data on every URL that has been

00:02:31.970 --> 00:02:35.550
submitted to the URLhaus that's been classified as Dridex.

00:02:35.550 --> 00:02:37.970
And that classification comes from a number of sources.

00:02:37.970 --> 00:02:42.250
It can be user submitted, so let's say you and I found some URLs right now.

00:02:42.260 --> 00:02:44.770
We knew they were dropping Dridex.

00:02:44.770 --> 00:02:46.660
We could tag that as we submitted.

00:02:46.660 --> 00:02:50.810
Abuse.ch also has its own process of identifying and

00:02:50.810 --> 00:02:53.480
adding those signatures as well, YARA rules,

00:02:53.480 --> 00:02:56.940
and IDS actually plays a pretty big role in that infrastructure.

00:02:56.950 --> 00:02:59.690
Once the response comes back, it's going to be JSON.

00:02:59.690 --> 00:03:03.590
We'll pipe that through a very handy command line utility called jq.

00:03:03.810 --> 00:03:08.700
If you haven't heard of it, jq will just help us to work with that JSON data.

00:03:08.700 --> 00:03:12.160
So we'll filter it, and we'll extract just those URLs.

00:03:12.160 --> 00:03:15.890
What this will end up looking like then‑‑‑here's some results.

00:03:15.890 --> 00:03:19.350
These are from the time of the recording anyway,

00:03:19.350 --> 00:03:23.170
the most thousand recent URLs that were submitted.

00:03:23.170 --> 00:03:25.890
The database goes back quite a bit further,

00:03:25.890 --> 00:03:28.310
and with their API being available,

00:03:28.320 --> 00:03:32.560
you can search for a specific domain rather than just consuming what

00:03:32.560 --> 00:03:35.840
has already been submitted in this broader list.

00:03:35.850 --> 00:03:38.810
So this is a really great place to go and look.

00:03:38.810 --> 00:03:42.000
I would say that if any of the requests in my

00:03:42.000 --> 00:03:44.650
environment are showing up on the URLhaus,

00:03:44.650 --> 00:03:46.260
there's probably an issue there.

00:03:46.260 --> 00:03:50.800
Not only that, but oftentimes the URLhaus and abuse.ch,

00:03:50.800 --> 00:03:51.870
they have other projects.

00:03:51.880 --> 00:03:56.070
They do a really good job of providing those signatures.

00:03:56.070 --> 00:03:58.150
You go to a place like VirusTotal,

00:03:58.160 --> 00:04:03.860
and you'll get 74 out of 87 antiviruses detect this as bad.

00:04:04.240 --> 00:04:09.000
Well, that's great, and it typically does better with files rather than URLs.

00:04:09.010 --> 00:04:10.060
‑Who do you trust?

00:04:10.070 --> 00:04:10.440
I don't know?

00:04:10.440 --> 00:04:14.080
Do I trust the 74 or do I trust the other, like, 12?

00:04:14.090 --> 00:04:15.360
I don't know.

00:04:15.840 --> 00:04:16.790
‑And what is it?

00:04:16.800 --> 00:04:21.470
That's the other thing is, how do you take 60 AD labels and distill it and say,

00:04:21.470 --> 00:04:23.900
oh, this is Dridex or, oh, this is malware.

00:04:23.910 --> 00:04:25.010
‑Well, it's bad.

00:04:25.010 --> 00:04:25.780
That's all we really know.

00:04:25.780 --> 00:04:26.700
‑That's all you really know.

00:04:26.700 --> 00:04:31.110
So it does that distillation for you, so it can really be quite powerful.

00:04:31.110 --> 00:04:34.390
So it has all of this data that's available.

00:04:34.400 --> 00:04:35.920
It also has rule sets,

00:04:35.920 --> 00:04:41.350
so we talked a lot about IDS now. You can get the the SSL malicious cert list,

00:04:41.350 --> 00:04:45.500
you can get IP lists, you can get lists that are,

00:04:45.500 --> 00:04:48.500
or rule sets that are made specifically for Suricata.

00:04:48.510 --> 00:04:51.350
And I really encourage you to check those out.

00:04:51.360 --> 00:04:52.950
Some of them are going to generate more noise,

00:04:52.950 --> 00:04:55.730
and so it's something that you have to test in your environment,

00:04:55.740 --> 00:04:57.700
maybe you don't deploy it across the entire enterprise.

00:04:57.700 --> 00:04:59.760
All those things have to be considered.

00:04:59.760 --> 00:05:02.800
The other thing that's interesting to point out here is you get a

00:05:02.800 --> 00:05:07.070
pretty good example as to the number of just,

00:05:07.080 --> 00:05:11.900
or the variety in the compromised hosts that Dridex is using.

00:05:11.910 --> 00:05:16.120
While I'm sure none of these sites I would recognize,

00:05:16.120 --> 00:05:20.240
as many of you probably wouldn't, they're all legitimate websites.

00:05:20.250 --> 00:05:23.840
In fact, if you look towards the top, there's logitech.pe,

00:05:23.840 --> 00:05:27.800
kind of a domain‑squatting sort of a scenario there.

00:05:27.800 --> 00:05:29.440
It's not all WordPress,

00:05:29.450 --> 00:05:34.980
although you can certainly see a lot of WordPress sites. Anything that has wp‑,

00:05:34.980 --> 00:05:37.410
it's pretty strong indicator that that's a WordPress.

00:05:37.420 --> 00:05:41.200
And then the resource, the file that's being requested,

00:05:41.210 --> 00:05:43.620
there's a lot of .php, there's some .tar,

00:05:43.620 --> 00:05:44.610
there's some .zip.

00:05:44.620 --> 00:05:48.750
All of that's actually irrelevant because the response is returning a PE file.

00:05:51.740 --> 00:05:51.930
‑Yeah.

00:05:51.930 --> 00:05:56.380
And we mentioned there's no real obfuscation but they do

00:05:56.380 --> 00:06:00.020
just variate the extension of the file,

00:06:00.020 --> 00:06:01.650
even though it doesn't actually change the file.

00:06:02.540 --> 00:06:03.270
‑Right, right.

00:06:03.270 --> 00:06:06.810
So you'll see a lot of times the PHP files and

00:06:06.810 --> 00:06:08.260
we'll even the .tar and the .zips,

00:06:08.260 --> 00:06:10.680
they just have this sort of random‑looking name.

00:06:10.690 --> 00:06:13.330
Something was generated in order to give those a name.

00:06:13.330 --> 00:06:16.870
How the site is structured, how are they were able to compromise that,

00:06:16.880 --> 00:06:18.990
what they ultimately have access to,

00:06:18.990 --> 00:06:23.250
that's going to also dictate to an extent the full URL that they use.

00:06:23.440 --> 00:06:27.680
Oftentimes, these hosts will serve multiple PE files,

00:06:27.680 --> 00:06:30.380
there'll be multiple locations that they're set up,

00:06:30.390 --> 00:06:31.450
different paths.

00:06:31.460 --> 00:06:36.590
So just because you get a payload from one URL,

00:06:36.590 --> 00:06:38.950
it doesn't mean that that same payload isn't six or

00:06:38.950 --> 00:06:40.460
seven other places on that host.

00:06:42.140 --> 00:06:42.920
‑Yeah, absolutely.

00:06:42.920 --> 00:06:45.730
And then so we take these URLs that we just got back,

00:06:45.730 --> 00:06:51.080
and now we're going to look for this activity historically in our environment.

00:06:51.090 --> 00:06:52.470
That the ide?

00:06:52.470 --> 00:06:54.430
‑Yeah, it can go either way.

00:06:54.440 --> 00:06:59.900
If we're specifically concerned about a Dridex infection,

00:06:59.910 --> 00:07:02.290
maybe some of the previous indicators that we've talked

00:07:02.290 --> 00:07:05.990
about are leading us down that path, we might pull these URLs and say,

00:07:05.990 --> 00:07:09.630
all righty, have any of these domains shown up in the environment?

00:07:09.630 --> 00:07:10.890
That would be one way to do it.

00:07:10.900 --> 00:07:15.480
We could take the URLs that we find and search for them on the URLhaus,

00:07:15.480 --> 00:07:18.820
and then if they show up we might get better fidelity than on,

00:07:18.820 --> 00:07:20.040
well, one, if they're even there.

00:07:20.040 --> 00:07:22.290
If they're there, that's an issue.

00:07:22.290 --> 00:07:25.050
And two, maybe get some classification.

00:07:25.050 --> 00:07:27.480
We'll see some historical context because the

00:07:27.480 --> 00:07:31.560
URLhaus provides not just the domain, but a lot of information about it.

00:07:31.560 --> 00:07:34.470
When was it reported, those signatures as we talked about,

00:07:34.470 --> 00:07:36.200
and other tags get applied to it.

00:07:36.210 --> 00:07:40.400
And so it can help to better understand maybe the

00:07:40.400 --> 00:07:42.710
broader context of the infection,

00:07:42.720 --> 00:07:46.560
as well as maybe more data and more fidelity on what's going on.

00:07:47.940 --> 00:07:51.380
‑And so we take these URLs, and these URLs are really what's

00:07:51.380 --> 00:07:54.740
inside that maldoc, so on it goes all the way full circle back

00:07:54.740 --> 00:07:56.750
to the maldocs in the first place.

00:07:57.140 --> 00:08:01.630
So you mentioned before we might take a look at what that script looks like

00:08:01.630 --> 00:08:05.570
and how it contains URLs that it's going to randomly select.

00:08:05.640 --> 00:08:06.760
Do you have an example of that?

00:08:07.540 --> 00:08:09.550
‑Yeah, so unfortunately,

00:08:09.550 --> 00:08:12.520
we don't have the time to get into the analysis of a maldoc,

00:08:12.530 --> 00:08:14.950
but that is something that I've spent a great deal of time on

00:08:14.950 --> 00:08:17.310
and worked on with some colleagues as well.

00:08:17.320 --> 00:08:23.590
An interesting thing with these Dridex documents is they embed a

00:08:23.590 --> 00:08:27.270
lot of domains into the maldocs themselves.

00:08:27.270 --> 00:08:30.730
And then when the maldoc executes, they decode,

00:08:30.730 --> 00:08:33.980
they obfuscate these domains, and there's, last I checked,

00:08:33.980 --> 00:08:37.159
four or five different techniques in which they do this obfuscation.

00:08:37.539 --> 00:08:39.860
They decode those and then they pick one at random,

00:08:39.860 --> 00:08:43.309
and so you could have a document that you execute once,

00:08:43.309 --> 00:08:47.080
and it tries to reach out to host1.xyz.

00:08:47.090 --> 00:08:50.230
You run it again and then it reaches out to host2.abc.

00:08:50.230 --> 00:08:52.440
It can vary.

00:08:52.440 --> 00:08:54.950
But because all of that content is there,

00:08:54.960 --> 00:08:57.090
if we can extract all those domains,

00:08:57.090 --> 00:09:02.830
now we can possibly obtain quite a bit better understanding

00:09:02.830 --> 00:09:05.670
as to other network‑based indicators.

00:09:05.680 --> 00:09:08.060
So you'll see this command.

00:09:08.060 --> 00:09:12.660
It's a Python script that's actually available on GitHub under BromiumLabs,

00:09:12.660 --> 00:09:15.360
and I'm sure we'll include a link in the course resources.

00:09:15.540 --> 00:09:21.080
This script was designed really to recreate the logic in

00:09:21.080 --> 00:09:26.500
the macro code to then parse the files, identify those indicators,

00:09:26.500 --> 00:09:29.230
and you'll see it's just going to print them out to a file.

00:09:29.240 --> 00:09:31.640
So this was one document,

00:09:31.650 --> 00:09:35.660
and at the very end of this command I just cat the contents.

00:09:35.670 --> 00:09:39.730
56 domains were extracted, so just from one single document.

00:09:39.730 --> 00:09:42.000
So from like a threat research perspective,

00:09:42.000 --> 00:09:43.250
the scale's really awesome,

00:09:43.440 --> 00:09:48.490
because now we can just harvest these Excel docs and scrape URLs

00:09:48.490 --> 00:09:50.540
and then submit them to the URLhaus all day long.

00:09:50.550 --> 00:09:52.340
From an incident perspective,

00:09:52.350 --> 00:09:55.010
this would be very valuable because I would want to look at

00:09:55.010 --> 00:09:58.500
this and take any one of these domains now,

00:09:58.500 --> 00:10:01.200
these URLs, and look for that in my environment, as well,

00:10:01.200 --> 00:10:03.360
because maybe the incident I'm investigating,

00:10:03.740 --> 00:10:07.030
two or three other people got the same document and it reached out to a

00:10:07.030 --> 00:10:09.880
different node and had slightly different behaviors.

00:10:09.880 --> 00:10:13.440
It was, it was a regsvr32 versus a RunDLL or something.

00:10:13.440 --> 00:10:16.580
‑We can start to assume, and maybe it's not true,

00:10:16.580 --> 00:10:18.940
but it's a great starting place to say that the email

00:10:18.940 --> 00:10:21.740
addresses have the same URL at the end,

00:10:21.740 --> 00:10:25.930
and if that was being sprayed out in kind of an automated sense that they

00:10:25.930 --> 00:10:29.560
likely got the same document within a specific period of time.

00:10:29.940 --> 00:10:32.890
And so excellent, yeah, from a behavioral perspective,

00:10:32.890 --> 00:10:35.850
even from how an attacker changes the malware perspective,

00:10:36.340 --> 00:10:39.350
it's reasonable to assume that at least more than one person,

00:10:39.350 --> 00:10:40.960
if not your entire organization,

00:10:40.960 --> 00:10:43.450
got hit with the same doc around the same period of time.

00:10:43.940 --> 00:10:44.590
‑Yeah, right,

00:10:44.600 --> 00:10:51.660
or slight variations of‑‑‑and this goes to highlight where not every

00:10:51.660 --> 00:10:56.810
organization has the budget to have a full security team to include

00:10:56.810 --> 00:10:59.110
malware analysts and threat researchers.

00:10:59.110 --> 00:11:03.430
But just understanding that maybe there's some tools out there like

00:11:03.430 --> 00:11:05.880
this script that can help for some quick wins,

00:11:05.880 --> 00:11:06.370
really.

00:11:06.380 --> 00:11:12.820
You don't have to get in and reverse engineer the maldoc or anything here.

00:11:12.820 --> 00:11:14.300
You're really dependent upon the tool.

00:11:14.300 --> 00:11:16.490
Of course, with that dependency comes, well,

00:11:16.490 --> 00:11:21.410
if this changes, and this changes all the time. As we develop these scripts,

00:11:21.410 --> 00:11:24.280
every couple of weeks they change something very

00:11:24.280 --> 00:11:26.300
minor in how the macro code executes,

00:11:26.310 --> 00:11:28.760
which then necessitates a change on our end.

00:11:28.940 --> 00:11:30.550
And so if you're using our tool,

00:11:30.550 --> 00:11:34.210
you become a bit dependent on us staying on top of things.

00:11:34.210 --> 00:11:37.880
But as long as that tool source is staying on top of things,

00:11:37.890 --> 00:11:39.720
it can be very valuable.

00:11:39.730 --> 00:11:47.410
And having some malware analysis capability can be really helpful because you

00:11:47.410 --> 00:11:52.890
may have never even known that this mountain of IOCs existed had you not been

00:11:52.890 --> 00:11:55.060
able to pull these documents apart a little bit.

00:11:55.640 --> 00:11:59.230
‑Yeah, and like you showed, a few seconds to grab the document,

00:11:59.230 --> 00:12:00.340
run this against it,

00:12:00.340 --> 00:12:03.990
and now that just really made your day from an incident response perspective,

00:12:03.990 --> 00:12:04.730
for sure.

00:12:04.730 --> 00:12:05.060
‑Yeah.

00:12:06.640 --> 00:12:07.630
‑So where do we go from here?

00:12:07.640 --> 00:12:08.890
What's what's our next step?

00:12:08.910 --> 00:12:11.370
So, we have these to look for,

00:12:11.370 --> 00:12:14.980
and now we're going to go look kind of for network indicators across

00:12:14.980 --> 00:12:17.430
the environment? Is that the next idea? ‑Yeah,

00:12:17.430 --> 00:12:19.280
so there's a few other,

00:12:19.280 --> 00:12:23.630
there's some other data that's generated that can be helpful in

00:12:23.630 --> 00:12:28.250
trying to observe this activity and broadening the scope in which

00:12:28.250 --> 00:12:30.460
we're doing some of that discovery.

00:12:30.940 --> 00:12:33.230
So we'll go back to to EveBox,

00:12:33.230 --> 00:12:35.760
we'll take a look at a couple of additional alerts.

00:12:35.760 --> 00:12:39.760
The two alerts we're going to focus on is the JA3 hash,

00:12:39.770 --> 00:12:45.290
which is generated by Suricata as it's processing the network traffic,

00:12:45.290 --> 00:12:47.470
particularly the TLS certificate information.

00:12:47.480 --> 00:12:51.410
And then this other alert, which is in yellow,

00:12:51.420 --> 00:12:55.230
which is the Executable Retrieved With Minimal HTTP Headers.

00:12:55.240 --> 00:12:59.970
So let's start with the minimal HTTP headers because what this is telling us,

00:12:59.970 --> 00:13:02.120
and when we talked about flowbits a little bit ago,

00:13:02.120 --> 00:13:04.960
and saying, well, rules,

00:13:05.240 --> 00:13:09.220
it's hard to make them understand broader context in

00:13:09.220 --> 00:13:11.890
which they're generating alerts on, the traffic that they're looking at.

00:13:11.890 --> 00:13:13.850
We can't really tie events together.

00:13:13.860 --> 00:13:16.520
We can, but it's it's really limited.

00:13:16.530 --> 00:13:22.530
And this is an example of that in that this minimal HTTP header is saying,

00:13:22.530 --> 00:13:22.750
all right,

00:13:22.750 --> 00:13:28.460
if I see an HTTP request that only has one or two or three HTTP headers,

00:13:28.740 --> 00:13:30.910
that's probably from some sort of a script,

00:13:30.920 --> 00:13:33.320
not from a user in their browser,

00:13:33.330 --> 00:13:36.180
and therefore, it's more likely to be suspicious.

00:13:36.180 --> 00:13:40.200
So let's set that flowbit, and then if in that same flow a PE file's downloaded,

00:13:40.200 --> 00:13:42.400
all right, now let's generate our alerts,

00:13:42.410 --> 00:13:46.620
and we hopefully have a little bit better fidelity on the situation.

00:13:46.620 --> 00:13:53.950
And that type of TTP, then, we can use that to alert on all sorts of activity,

00:13:53.950 --> 00:13:59.630
not just that minimal HTTP header, but really broadens the capability for

00:13:59.630 --> 00:14:02.850
not only prevention and detection but even response.

00:14:02.850 --> 00:14:04.540
Because another important thing,

00:14:04.540 --> 00:14:07.820
a lot of folks think about IDS alerts as this sort of forward‑moving thing,

00:14:07.820 --> 00:14:09.560
they're being generated in real time.

00:14:09.940 --> 00:14:12.590
But if you have archival network traffic,

00:14:12.590 --> 00:14:16.850
you can pull swaths of your traffic, run it through Suricata in offline mode,

00:14:16.850 --> 00:14:18.350
and it'll go back and generate IDS alerts.

00:14:18.430 --> 00:14:21.010
So let's say you're performing an investigation,

00:14:21.020 --> 00:14:22.960
you've crafted this new rule,

00:14:22.970 --> 00:14:28.630
now you can apply it to Suricata and go back and run a pcap from 24 hours ago.

00:14:28.630 --> 00:14:31.040
And it'll generate new alerts based off of those

00:14:31.050 --> 00:14:32.360
updated rules that you're using.

00:14:33.040 --> 00:14:36.560
‑Yeah, that's amazing, and applause to you.

00:14:36.560 --> 00:14:38.860
Great callout on the HTTP headers.

00:14:38.860 --> 00:14:42.260
From perspective detection and behavioral detections,

00:14:42.260 --> 00:14:44.590
like I said before, it's a big keyword for me.

00:14:44.590 --> 00:14:49.820
What a really cool artifact of things being automated in

00:14:49.820 --> 00:14:52.700
the way that that malware communicates, is to look at minimal HTTP headers.

00:14:52.700 --> 00:14:54.320
That's a great callout.

00:14:54.330 --> 00:14:55.240
Probably should put a note.

00:14:55.240 --> 00:14:57.960
I'm sure we can find some sort of good writeup on that too.

00:14:58.340 --> 00:15:01.100
Probably should put a note for that in the exercise follows as well,

00:15:01.110 --> 00:15:03.880
because, like you said, that's more than just Dridex.

00:15:03.890 --> 00:15:06.090
That's something we can use to chain to other stuff,

00:15:06.090 --> 00:15:09.770
and I think is really, really helpful from a defender perspective,

00:15:09.780 --> 00:15:11.930
start getting in that same kind of mindset.

00:15:11.940 --> 00:15:15.550
That's one I hadn't‑‑I don't think I've actually used that one before.

00:15:15.550 --> 00:15:17.260
So that's one that's excellent.

00:15:17.740 --> 00:15:18.250
Thank you for that.

00:15:19.140 --> 00:15:20.650
‑Yeah, so to help with that,

00:15:20.650 --> 00:15:24.280
we're going to transition to one of my favorite tools, Arkime,

00:15:24.280 --> 00:15:25.920
formerly known as Moloch.

00:15:25.930 --> 00:15:34.500
Arkime is amazing in that it really makes it easy to look and hunt

00:15:34.500 --> 00:15:36.640
and pivot through your network traffic, because it essentially

00:15:36.640 --> 00:15:39.920
takes all of that network traffic data, particularly the metadata,

00:15:39.920 --> 00:15:43.880
and it makes it a facet, it makes it so that we can search.

00:15:43.880 --> 00:15:47.550
So I have a full course on Arkime if you're interested in that.

00:15:47.560 --> 00:15:50.080
Again, we don't have time to get in all the details here.

00:15:50.080 --> 00:15:55.830
But what we're looking at is this breakdown of HTTP‑specific data,

00:15:55.830 --> 00:15:59.070
and not only can we see very specific data,

00:15:59.080 --> 00:16:02.640
like the value from the hostname header or the user agent,

00:16:02.650 --> 00:16:07.420
but then it also has the ability to interpret that data for us,

00:16:07.420 --> 00:16:10.790
to say, well, let's look at the Has Src Header count.

00:16:10.790 --> 00:16:15.940
And now using that property, we can create a filter,

00:16:15.950 --> 00:16:17.400
and we could say, all right,

00:16:17.400 --> 00:16:23.340
let's look for any HTTP request that has a source header count less than 4.

00:16:23.350 --> 00:16:26.140
And now we've taken that rule,

00:16:26.150 --> 00:16:29.560
maybe that was only looking for 2 or less and we've

00:16:29.560 --> 00:16:33.710
made it a little bit broader, or we've made it certainly flexible.

00:16:33.710 --> 00:16:36.000
Because now we're not writing rules at this point;

00:16:36.000 --> 00:16:40.270
we're just exploring our traffic, and we're seeing the results that come back.

00:16:40.270 --> 00:16:42.880
We're taking this this mountain of network data,

00:16:42.890 --> 00:16:44.660
and we're starting to narrow it down.

00:16:44.670 --> 00:16:47.820
Here we can see we have two requests.

00:16:47.830 --> 00:16:51.560
Both of these requests are to a suspicious‑looking host,

00:16:51.560 --> 00:16:55.650
a not very common TLD, a PDF file.

00:16:56.340 --> 00:17:01.520
It's requested over port 80, and if we were to expand that session,

00:17:01.530 --> 00:17:06.550
there's a PE file, and that, of course, would raise our suspicions even further.

00:17:06.560 --> 00:17:12.490
And so now we've been able to take these broader techniques that we see

00:17:12.490 --> 00:17:17.109
with how this malware operates and use that to turn that into a proactive

00:17:17.109 --> 00:17:20.140
hunting technique to help us in kind of scoping,

00:17:20.140 --> 00:17:23.000
or were there other hosts? I mean, that's probably the biggest goal here,

00:17:23.010 --> 00:17:28.250
is how many‑‑‑can we identify maybe previously undetected malicious behavior,

00:17:28.250 --> 00:17:30.670
or if we're in like that incident response mode,

00:17:30.670 --> 00:17:34.200
are there other hosts that are part of this attack?

00:17:34.200 --> 00:17:37.310
And we need to get them into scope and investigate and isolate and do all

00:17:37.310 --> 00:17:40.050
the things that we would do in our incident response plan.

00:17:41.340 --> 00:17:43.770
‑Right, and we're pattern matching at this point.

00:17:43.780 --> 00:17:46.670
And when you're in the incident response flex motion,

00:17:46.670 --> 00:17:49.780
we're pattern matching for our specific environment for the

00:17:49.780 --> 00:17:51.300
infection in our specific environment.

00:17:51.300 --> 00:17:54.860
So we got URLs, but we also have, again,

00:17:54.870 --> 00:17:58.600
awesome callout on headers and not just using that as a behavior like,

00:17:58.600 --> 00:18:00.580
oh, it seems like there's few HTTP headers.

00:18:00.580 --> 00:18:04.620
More of, like, in this specific event that we're concerned about,

00:18:04.630 --> 00:18:05.990
there's less than four.

00:18:06.000 --> 00:18:08.060
And when I look across all of my network traffic,

00:18:08.440 --> 00:18:11.710
that reduces that down to two sessions, and that should tell me something,

00:18:11.720 --> 00:18:16.290
is we're taking behavioral indicators and using them as signatures.

00:18:16.300 --> 00:18:19.630
And so, yeah, of course, we don't‑‑‑signatures versus behavioral indicators.

00:18:19.630 --> 00:18:24.920
But still the behavior of less than four headers ends up being something we can

00:18:24.920 --> 00:18:28.350
look for across our environment near all that traffic down,

00:18:28.350 --> 00:18:31.570
like you said, identify their hosts that are exhibiting the same behavior.

00:18:31.570 --> 00:18:33.160
And that helps us scope.

00:18:33.170 --> 00:18:34.530
Now, we know what we need to do.

00:18:34.530 --> 00:18:36.710
We can ring the bell and say, hey,

00:18:36.720 --> 00:18:40.270
let's remediate all these devices and make some edge

00:18:40.270 --> 00:18:41.590
rules to block this kind of traffic,

00:18:41.590 --> 00:18:44.370
and then I can go home and I can watch my Disney+

00:18:44.370 --> 00:18:45.650
show that just came out on Friday.

00:18:45.660 --> 00:18:50.310
‑Those are the type of patterns that we can look for that can be

00:18:50.310 --> 00:18:55.770
really powerful in trying to identify this suspicious,

00:18:55.770 --> 00:18:56.840
malicious activity.

00:18:56.850 --> 00:19:01.490
Another area that I'd like to focus on, then, is the JA3 hashes.

00:19:01.500 --> 00:19:02.550
‑Yeah, please do.

00:19:02.550 --> 00:19:03.660
Also, thanks, Salesforce.

00:19:03.660 --> 00:19:05.130
And I want to make a quick callout.

00:19:05.130 --> 00:19:07.900
Arkime is from AOL, just in case.

00:19:07.910 --> 00:19:08.880
So thanks, AOL.

00:19:08.880 --> 00:19:10.510
Lots of the time put into both these things that

00:19:10.510 --> 00:19:12.050
are really helpful for all of us.

00:19:13.440 --> 00:19:14.900
‑Yes, absolutely.

00:19:14.910 --> 00:19:20.710
So sticking within Arkime, there is in the same view,

00:19:20.710 --> 00:19:22.710
and Arkime has different views that we can look at the

00:19:22.710 --> 00:19:25.870
data, this view's breaking things down by protocol, we

00:19:25.870 --> 00:19:29.270
have the TLS. And this allows us, then,

00:19:29.270 --> 00:19:33.470
to see not only just the raw data from the TLS session,

00:19:33.470 --> 00:19:35.530
the certificate that went back and forth,

00:19:35.540 --> 00:19:38.110
but then that data that it generated, the JA3 hash,

00:19:38.110 --> 00:19:39.110
in this case.

00:19:39.110 --> 00:19:43.540
JA3 hashes, again, it's not a precise science.

00:19:43.540 --> 00:19:47.620
It does lend itself to false positives based off of the

00:19:47.620 --> 00:19:50.170
way that this hash is being generated.

00:19:50.340 --> 00:19:52.640
Previously, we were looking at HTTP,

00:19:52.640 --> 00:19:57.240
and the reason that we could see the HTTP data is because

00:19:57.240 --> 00:19:58.650
it wasn't over an encrypted session.

00:19:58.660 --> 00:20:02.670
So either the connection just doesn't use TLS,

00:20:02.680 --> 00:20:06.810
or maybe the organization is able to pry into that

00:20:06.810 --> 00:20:09.790
secure connection to see that content.

00:20:09.790 --> 00:20:12.060
Not as common as a scenario.

00:20:12.060 --> 00:20:14.850
So threat actors that use a TLS session,

00:20:14.850 --> 00:20:17.320
we lose visibility, but we don't lose total visibility.

00:20:17.320 --> 00:20:19.060
We still have some information,

00:20:19.080 --> 00:20:23.950
such as the hosts that they're connecting to and information from the TLS cert,

00:20:23.950 --> 00:20:26.090
as well as things that can be generated.

00:20:26.100 --> 00:20:28.410
So that's where the JA3 hash comes in.

00:20:28.420 --> 00:20:32.750
It's generated based off of some of the data from the TLS certificate,

00:20:32.750 --> 00:20:35.600
and it's another open source technology that you can go

00:20:35.600 --> 00:20:37.510
to the GitHub and read all about it.

00:20:37.520 --> 00:20:42.590
This hash, so any TLS session that is being recorded‑‑‑in this case,

00:20:42.590 --> 00:20:46.420
the traffic is either being recorded by Arkime so it can do

00:20:46.420 --> 00:20:48.240
the network capture for you on the network,

00:20:48.250 --> 00:20:51.670
or it can also do pcap replay.

00:20:51.940 --> 00:20:54.270
So it's going to generate those hashes.

00:20:54.270 --> 00:20:58.520
And once those hashes are there, we can we can pivot on them.

00:20:58.520 --> 00:20:58.920
So,

00:20:58.930 --> 00:21:01.670
earlier we talked about the file hash and that if we took the

00:21:01.670 --> 00:21:05.910
file hash and we filtered by the Suricata data,

00:21:05.920 --> 00:21:07.480
it has that file information.

00:21:07.480 --> 00:21:11.040
So any session that has that file hash, we can filter on that.

00:21:11.040 --> 00:21:13.060
And then it's very similar here with the JA3.

00:21:13.070 --> 00:21:16.810
We can now create a filter that says any session that

00:21:16.810 --> 00:21:20.700
matches this generated JA3 hash, let me see it.

00:21:20.710 --> 00:21:23.760
So here we have, what do we have?

00:21:23.760 --> 00:21:27.430
Two unique sessions, two different hosts,

00:21:27.440 --> 00:21:30.040
and they're communicating with the same hash,

00:21:30.440 --> 00:21:34.870
but they're also over non‑standard and both different TLS ports.

00:21:34.880 --> 00:21:39.910
And so maybe we had initially some observed bad

00:21:39.910 --> 00:21:41.410
activity from one of these hosts.

00:21:41.420 --> 00:21:44.430
Well, this JA3 would possibly allow us, then,

00:21:44.430 --> 00:21:47.080
to pivot and get this other host into purview.

00:21:47.080 --> 00:21:49.540
And certainly we would want to make that part of our investigation.

00:21:49.540 --> 00:21:54.320
‑For Dridex specifically, the initial piece of this was,

00:21:54.320 --> 00:21:58.170
it reached out and we see that behavior being over HTTP,

00:21:58.180 --> 00:21:59.290
or not encrypted.

00:21:59.290 --> 00:22:02.000
It's not using TLS to pull down the executable.

00:22:02.010 --> 00:22:03.760
And that was kind of like that detection flow.

00:22:04.240 --> 00:22:08.110
Now, if you have an active infection, the C2,

00:22:08.110 --> 00:22:12.160
because we did do the certificate alert as well, that C2,

00:22:12.160 --> 00:22:13.720
that's using TLS.

00:22:13.730 --> 00:22:14.960
Is that true?

00:22:15.440 --> 00:22:15.540
‑Right, yes.

00:22:15.540 --> 00:22:19.890
‑And so now this allows us to then do, hey,

00:22:19.900 --> 00:22:21.310
the first phase of attack, like,

00:22:21.310 --> 00:22:25.460
if you saw these set of IPs or endpoints that alerted

00:22:25.460 --> 00:22:29.600
on the HTTP with the PE download, then they got to that stage.

00:22:29.600 --> 00:22:32.340
If we missed all of that and we just see C2,

00:22:32.340 --> 00:22:34.770
it still means that probably happened in the past,

00:22:34.770 --> 00:22:36.840
and we now have an active infection.

00:22:36.840 --> 00:22:37.320
‑Right.

00:22:37.320 --> 00:22:40.250
‑So really cool how we can use these different kinds of

00:22:40.250 --> 00:22:44.130
detections to focus on different parts of even the specific

00:22:44.130 --> 00:22:45.670
malwares like chain of compromise.

00:22:47.040 --> 00:22:47.540
‑Yeah, yeah.

00:22:47.540 --> 00:22:51.060
And and it's very common that the patterns to bring this back to

00:22:51.060 --> 00:22:54.020
being a little bit more about Dridex again,

00:22:54.020 --> 00:22:57.890
the C2, the command and control after infection,

00:22:57.890 --> 00:22:59.690
very common for it to use TLS.

00:22:59.690 --> 00:23:01.860
A number of different hosts.

00:23:01.870 --> 00:23:05.710
Oftentimes it's going to communicate with those not via DNS.

00:23:05.710 --> 00:23:08.560
It's not going to have to resolve the hostname; it's

00:23:08.560 --> 00:23:11.330
just going to direct to an IP address.

00:23:11.340 --> 00:23:16.580
It's also common to see these nonstandard ports as well.

00:23:16.590 --> 00:23:22.040
And so another view on our network traffic would be to say,

00:23:22.050 --> 00:23:26.300
show me any TLS session that is not to port 443,

00:23:26.310 --> 00:23:30.500
and that might elicit something worth inspecting.

00:23:30.500 --> 00:23:32.950
And then you could cross‑reference those IPs with some of the other

00:23:32.950 --> 00:23:35.950
techniques that we've been discussing and help to understand.

00:23:36.140 --> 00:23:39.300
It's sort of a blurry line we get into and then when you start

00:23:39.300 --> 00:23:43.710
trying to really respond to an infection you kind of go into

00:23:43.710 --> 00:23:45.520
this proactive threat‑hunting mode,

00:23:45.530 --> 00:23:49.640
and maybe you're going to start uncovering other things you didn't know about.

00:23:49.650 --> 00:23:54.580
And so you may take a note of that something is abnormal here,

00:23:54.580 --> 00:23:57.370
but keep in mind the task at hand,

00:23:57.370 --> 00:24:01.340
which is this infection that now we're trying to to really uncover,

00:24:01.340 --> 00:24:05.850
unearth, scope, and begin mitigation and recovery.

00:24:06.640 --> 00:24:07.570
‑Where do we go from here?

00:24:07.570 --> 00:24:10.190
We have a really good, honestly, inspection,

00:24:10.190 --> 00:24:13.050
pretty much, of every phase of this kill chain.

00:24:13.640 --> 00:24:14.660
What are we missing?

00:24:16.040 --> 00:24:18.990
There's just a couple of things left that I'd like to point out.

00:24:19.000 --> 00:24:20.980
There's many different rabbit holes,

00:24:20.980 --> 00:24:23.810
so I'm not trying to say these are the only things you have to look at,

00:24:23.820 --> 00:24:25.360
but to me these are the big ones.

00:24:25.370 --> 00:24:29.730
We can expand that network traffic in Arkime.

00:24:30.120 --> 00:24:34.220
You just expand the session, and you'll see that more granular data,

00:24:34.220 --> 00:24:35.020
just like with alerts.

00:24:35.020 --> 00:24:36.530
We were drilling down with alerts.

00:24:36.540 --> 00:24:39.010
So here we have information about the TLS.

00:24:39.010 --> 00:24:41.320
Of course, we have the JA3 and JA3S.

00:24:41.320 --> 00:24:43.520
We also have information about the cert.

00:24:43.530 --> 00:24:45.070
So when was it issued?

00:24:45.070 --> 00:24:47.460
When does it expire?

00:24:47.460 --> 00:24:50.760
What is the subject organization, the subject common name?

00:24:51.140 --> 00:24:55.140
And that also becomes information that can help identify

00:24:55.140 --> 00:24:58.840
that a TLS session may be a bit suspicious,

00:24:58.840 --> 00:25:00.480
because here is an example.

00:25:00.480 --> 00:25:02.200
This is from Dridex, of course.

00:25:02.210 --> 00:25:07.670
And those are really uncommon names for the subject common and the subject org.

00:25:07.940 --> 00:25:12.740
And so creating signatures on that or using that to try to

00:25:12.750 --> 00:25:16.140
pivot in the network data can yield results.

00:25:16.140 --> 00:25:18.050
It's not always, but it certainly can.

00:25:18.060 --> 00:25:22.290
And it's another piece of data, particularly with the TLS session,

00:25:22.290 --> 00:25:27.170
that we can investigate to help us better understand what just occurred.

00:25:28.540 --> 00:25:33.200
‑Yeah, and dare I say you could take that and use it in a Suricata rule,

00:25:33.200 --> 00:25:37.690
like the rule that you showed us before in detection and kind of make

00:25:37.690 --> 00:25:40.050
something that matches that certificate to go a bit faster.

00:25:40.050 --> 00:25:42.060
Because we're looking at session data here,

00:25:42.060 --> 00:25:45.350
and being able to filter by that's one thing; having a really

00:25:45.350 --> 00:25:48.610
quick kind of PCRE detection can help too,

00:25:48.610 --> 00:25:50.550
especially in processing a big mountain of data.

00:25:51.540 --> 00:25:53.250
‑Right, yep, yep, absolutely.

00:25:53.260 --> 00:25:58.750
And, again, even stepping back, again, for a moment from Dridex specifically,

00:25:58.760 --> 00:26:02.960
you can look for just misconfiguration, OpenSSL,

00:26:02.960 --> 00:26:05.140
default generated TLS certificates.

00:26:05.150 --> 00:26:10.460
If I see Internet Widgits Pty, well, maybe that's not bad guys,

00:26:10.540 --> 00:26:13.210
but maybe that's some self‑signed certificate in the

00:26:13.210 --> 00:26:14.620
environment that shouldn't be there.

00:26:14.700 --> 00:26:20.350
So there's still a lot of value in not just responding,

00:26:20.350 --> 00:26:21.450
detecting threats,

00:26:21.450 --> 00:26:24.950
but even in just trying to figure out what's going on in your networks.

00:26:25.440 --> 00:26:28.950
The last thing is just a config extractor.

00:26:29.940 --> 00:26:35.720
So a lot of malware, again, if you think about it in terms of, its software.

00:26:35.720 --> 00:26:42.160
It's just the intent that changes, big picture. Like a lot of software,

00:26:42.160 --> 00:26:43.900
it'll have configuration files.

00:26:43.900 --> 00:26:49.240
And what what a lot of researchers and companies have gotten really good at

00:26:49.250 --> 00:26:53.560
over the last few years is identifying those families that have

00:26:53.560 --> 00:26:56.670
configurations and then figuring out how to extract them.

00:26:56.680 --> 00:26:58.270
So Triage,

00:26:58.280 --> 00:27:01.990
which is from more or less the same group that brought you the Cuckoo Sandbox,

00:27:01.990 --> 00:27:06.550
they do a phenomenal job of writing these config extractors.

00:27:06.550 --> 00:27:09.530
And so here you can see one of our samples from earlier today.

00:27:09.530 --> 00:27:12.340
We have the tagging that runs in those sort of

00:27:12.340 --> 00:27:15.040
color‑coded tags right underneath the score.

00:27:15.050 --> 00:27:17.110
They did a great job of identifying this as Dridex,

00:27:17.110 --> 00:27:20.210
and because they did that, they also have a config extractor.

00:27:20.220 --> 00:27:23.910
So, now they've extracted all of the C2 notes.

00:27:23.920 --> 00:27:27.320
And so with that information, we can, again,

00:27:27.320 --> 00:27:31.500
take that back to the network, look for any communication to those IPs,

00:27:31.510 --> 00:27:36.400
update our rules to block, whatever we need to do,

00:27:36.400 --> 00:27:39.720
and it helps because it broadens that approach.

00:27:39.720 --> 00:27:42.600
Just like with the Excel documents we talked about earlier,

00:27:42.610 --> 00:27:46.240
there's 56 domains in that one document that we looked at.

00:27:46.250 --> 00:27:47.780
And if we extract those,

00:27:47.790 --> 00:27:52.430
we can be much more thorough in knowing where that document communicated to.

00:27:52.440 --> 00:27:53.570
Same with the Trojan.

00:27:53.570 --> 00:27:55.370
Now we can be much more thorough in knowing,

00:27:55.380 --> 00:27:57.460
well, based on the configuration,

00:27:57.740 --> 00:28:00.330
it has these three posts that it's going to communicate to,

00:28:00.330 --> 00:28:01.960
so let's see if there's any traffic going there.

00:28:03.940 --> 00:28:06.670
‑I haven't seen this before, so thank you for showing it to me.

00:28:07.140 --> 00:28:08.270
Really, really cool.

00:28:08.640 --> 00:28:12.020
Maybe a lot of other people have seen it, but just awesome.

00:28:12.020 --> 00:28:15.790
Exactly what you said, we're in our incident response scenario,

00:28:15.790 --> 00:28:18.680
and it would be great to have the information that

00:28:18.680 --> 00:28:20.890
the C2's using for its config, and, here we go,

00:28:20.890 --> 00:28:22.380
we have a tool to be able to do that.

00:28:22.390 --> 00:28:25.030
That's going to make our day a lot easier.

00:28:25.040 --> 00:28:26.160
A really great thing.

00:28:26.160 --> 00:28:30.430
Again, we'll include some notes for this in the exercise files,

00:28:30.440 --> 00:28:35.360
so this is phenomenal, great information to pivot off of.

00:28:35.940 --> 00:28:41.120
‑Of course not every family has a config extractor created.

00:28:41.130 --> 00:28:44.990
They're just kind of strewn throughout the internet as well.

00:28:45.000 --> 00:28:49.130
But places like Cuckoo or, I'm sorry, places like Triage,

00:28:49.130 --> 00:28:52.410
probably the most popular Cuckoo forks, CAPE,

00:28:53.640 --> 00:29:00.250
they do a lot more with the config extractors. And much like

00:29:00.250 --> 00:29:01.950
the conversation around the Excel docs,

00:29:01.960 --> 00:29:08.050
it allows us to leverage some some really high‑level reverse engineering

00:29:08.050 --> 00:29:12.600
skills without having to do it all ourselves. Because if we had to extract

00:29:12.600 --> 00:29:18.420
these IPs, we'd either have to get lucky in the sense that if we throw it in

00:29:18.420 --> 00:29:22.250
a sandbox that it's going to try to communicate with all three of them. Not

00:29:22.250 --> 00:29:26.910
everything's going to do that. Or we have to have the time and the skills to

00:29:26.910 --> 00:29:31.770
get into and do that extraction ourselves, which can be really time

00:29:31.770 --> 00:29:32.370
consuming.

00:29:32.380 --> 00:29:35.930
You could spend weeks just trying to reverse engineer something. In that

00:29:35.930 --> 00:29:41.900
case, well, the damage is done. ‑The passwords are gone, man.

00:29:41.910 --> 00:29:43.810
‑Yeah.

00:29:43.820 --> 00:29:45.080
‑Leave them behind.

00:29:45.090 --> 00:29:50.550
Yeah, yeah, don't look at your bank account because it's not going to be great.

00:29:51.640 --> 00:29:54.100
So our focus here getting through the incident response,

00:29:54.110 --> 00:29:59.360
a ton of great tools to really, how do I detect it, how do I scope it, if I do

00:29:59.360 --> 00:30:03.440
detect it? So now I'm coming in and understanding every device that it's on.

00:30:03.440 --> 00:30:07.440
Ideally, if we're looking at initial access and I'm trying to understand root

00:30:07.440 --> 00:30:13.600
cause but really looking at phishing, likely a maldoc, and so we can stop that

00:30:13.600 --> 00:30:17.490
capability by blocking maybe just that one file but maybe also blocking some of

00:30:17.490 --> 00:30:18.500
these behaviors?

00:30:18.690 --> 00:30:20.760
And then clean up all those devices.

00:30:20.760 --> 00:30:22.390
So hopefully if you're responding to this,

00:30:22.390 --> 00:30:24.660
you have some sort of automated SOAR capability,

00:30:24.670 --> 00:30:28.990
you can click the button and it remediates, or your flashing drives,

00:30:28.990 --> 00:30:29.560
that kind of thing.

00:30:29.560 --> 00:30:34.350
But if you back up your user data. So really, really cool. Anything else you

00:30:34.350 --> 00:30:36.520
wanted to add to that piece for the incident response?

00:30:36.530 --> 00:30:42.460
Anything kind of left behind? Even recovery we can check for C2 with

00:30:42.470 --> 00:30:49.040
that config. ‑Yeah, with this type of threat actor,

00:30:49.040 --> 00:30:52.750
it's going to create noise on the network. It's

00:30:52.750 --> 00:30:54.690
going to create noise on the host.

00:30:54.700 --> 00:31:00.120
The modular behavior makes what happens on the host a little less predictable.

00:31:00.130 --> 00:31:03.460
What happens on the network, I think, is a little more predictable,

00:31:03.460 --> 00:31:06.800
although you could argue both ways.

00:31:06.810 --> 00:31:11.490
So that's really where a lot of this focus has been. It is going to communicate

00:31:11.490 --> 00:31:14.270
out; it does have a C2 network. That is how it's designed.

00:31:14.640 --> 00:31:18.570
Of course, that can change, as all malware families do. They're just like

00:31:18.570 --> 00:31:24.430
regular software. They evolve; they adapt. It's a constant back and forth with

00:31:24.440 --> 00:31:28.320
understanding what defenders are doing and just tweaking things enough to get

00:31:28.330 --> 00:31:31.460
past their detections and their capabilities.

00:31:31.460 --> 00:31:37.410
But it's a great place to start. I think there's a lot of, fortunately for us,

00:31:37.410 --> 00:31:41.830
a lot of strong indicators to build off of some really good research

00:31:41.830 --> 00:31:47.780
that's being done. The script I mentioned with the maldocs, a colleague

00:31:47.780 --> 00:31:51.000
is running that every day on all of the Excel,

00:31:51.010 --> 00:31:51.650
or I'm sorry,

00:31:51.650 --> 00:31:55.970
the Dridex documents that he gets his hands on and submitting that all to

00:31:55.970 --> 00:32:00.060
URLhaus. And so there's some great communities out there. Much like we

00:32:00.060 --> 00:32:03.930
saw with Cryptolaemus and Emotet, they're really focused on trying to

00:32:03.930 --> 00:32:09.180
take that group down or reduce its operational capability and let law

00:32:09.180 --> 00:32:10.440
enforcement take them down.

00:32:10.450 --> 00:32:14.530
A lot of support around this, since it is one of the

00:32:14.540 --> 00:32:17.050
more prevalent groups that's out there.

00:32:18.140 --> 00:32:18.480
‑Yeah,

00:32:18.490 --> 00:32:21.870
and that's always the idea. Even though like I can't go

00:32:21.870 --> 00:32:25.670
physically arrest the person who launched the phishing

00:32:25.680 --> 00:32:28.050
malware myself, but like you said,

00:32:28.050 --> 00:32:31.740
like we saw with Emotet, that will happen eventually. It just will. And the

00:32:31.740 --> 00:32:35.100
other piece of this is, in the meantime we can all work as a community with

00:32:35.100 --> 00:32:39.670
things like URLhaus and other open source indicators. There's plenty of

00:32:39.670 --> 00:32:43.460
people doing research, where we can make it more and more and more

00:32:43.640 --> 00:32:49.390
expensive for the actors behind Dridex to operate, and expensive being

00:32:49.390 --> 00:32:52.260
whether it's man hours, whether it's no return on investment,

00:32:52.260 --> 00:32:54.210
whether it's actual costs, because these people get paid

00:32:54.210 --> 00:32:58.600
too, or whatever it is, at some point, just like any other business,

00:32:58.610 --> 00:33:00.220
there's a, hey,

00:33:00.220 --> 00:33:04.190
I'm spending more money that I'm making. This really isn't worth it. And

00:33:04.190 --> 00:33:07.850
the more we can do that across the board, but again,

00:33:07.850 --> 00:33:09.360
this one's specifically about Dridex,

00:33:09.740 --> 00:33:12.990
the less likely they're going to be able to continue to innovate quickly.

00:33:13.000 --> 00:33:16.520
If they're making tons of money, they can pay people and completely outpace us.

00:33:16.520 --> 00:33:18.830
They can be a really big dev shop.

00:33:18.840 --> 00:33:20.430
But if they're not making any money and they can't pay

00:33:20.430 --> 00:33:23.350
people, we start to get into how we're impacting their

00:33:23.360 --> 00:33:25.360
business operations capability, like you said.
