TRANSKRYPCJA 🔗
WEBVTT
00:00:02.000 --> 00:00:10.320
Good morning, good afternoon, or good evening, whenever you may be, and welcome back to another episode of The SOC Brief.
00:00:10.480 --> 00:00:18.320
I'm your host Andrew, and today we're talking about a cybersecurity incident that's unfolding as we record this.
00:00:18.719 --> 00:00:32.079
Yesterday morning, Vanderbilt University in Nashville announced that it was responding to an active cybersecurity incident affecting university devices, systems, and potentially cloud-based services.
00:00:32.320 --> 00:00:37.679
And Vanderbilt's instructions to students, faculty, and staff were pretty direct.
00:00:38.000 --> 00:00:54.799
Shut down university devices, disconnect from university Wi-Fi, don't click anything suspicious, don't try to troubleshoot the problem yourself, and if something strange happens on your computer, turn it off and call IT from another device.
00:00:55.200 --> 00:01:02.719
That's a pretty significant response for a university with roughly 14,000 students and thousands of faculty and staff.
00:01:02.880 --> 00:01:13.280
And it gives us a really interesting incident response lesson, because sometimes the right security decision is also an incredibly disruptive business decision.
00:01:13.599 --> 00:01:17.200
Sometimes you really do have to pull the plug.
00:01:17.519 --> 00:01:20.719
So let's start with what we actually know.
00:01:21.120 --> 00:01:28.480
On Monday, November 9th, Vanderbilt University publicly confirmed that it was dealing with an active cybersecurity incident.
00:01:28.719 --> 00:01:35.359
The university warned that devices, internal systems, and cloud-based services could potentially be affected.
00:01:35.680 --> 00:01:57.439
Users were told to watch for things like unexpected pop-ups, security warnings, files being moved, renamed, or deleted, changes to files or permissions inside OneDrive, Box, or other cloud storage platforms, unfamiliar applications or prompts, and unexpected requests for passwords or MFA approvals.
00:01:57.680 --> 00:02:04.719
And Vanderbilt specifically told users if you see anything suspicious, don't investigate it.
00:02:04.879 --> 00:02:09.280
Don't click around, don't try to fix it, just shut the machine down.
00:02:09.520 --> 00:02:18.960
That may sound extreme, but during an active ransomware event, somebody trying to be helpful can sometimes make things significantly worse.
00:02:19.199 --> 00:02:24.319
Now, there's another name attached to this attack, BYOD.
00:02:24.719 --> 00:02:27.439
And I'm not talking about bring your own device.
00:02:27.680 --> 00:02:32.960
BYOD is also the name of a relatively new ransomware and extortion group.
00:02:33.120 --> 00:02:36.800
The group has claimed responsibility for the Vanderbilt incident.
00:02:37.039 --> 00:02:48.800
Screenshots provided to journalists show ransomware notes on Vanderbilt systems referencing BYOD, and some users reported files being renamed with a.byod extension.
00:02:49.039 -->
Load full transcript
00:03:00.240
There are also reports of compromised computers actually playing an audio message announcing that the network has been compromised, which I think is unnecessarily theatrical.
00:03:00.400 --> 00:03:06.240
Because apparently ransom nodes weren't dramatic enough anymore, we now need ransomware voice announcements.
00:03:06.560 --> 00:03:14.879
BYOD has claimed that it encrypted Vanderbilt databases and stole large amounts of student, employee, and research information.
00:03:15.120 --> 00:03:28.159
The group specifically claims to possess more than 1 million Social Security numbers along with names, email addresses, phone numbers, addresses, roles, birthdays, and other information.
00:03:28.639 --> 00:03:31.039
But here's a really important part.
00:03:31.599 --> 00:03:35.039
Those claims have not been confirmed by Vanderbilt.
00:03:35.199 --> 00:03:51.919
As of the time I'm recording this, but has not publicly confirmed that BYOD stole 1 million Social Security numbers, confirmed the attacker's claims about backups, or publicly provided a complete technical explanation of how the attackers initially gain access.
00:03:52.159 --> 00:03:56.400
And that's important because threat actors are not reliable narrators.
00:03:56.639 --> 00:04:01.919
Ransomware groups have every incentive to exaggerate the amount of access they have.
00:04:02.159 --> 00:04:08.719
Their entire business model depends on convincing the victim that the situation is as terrible as possible.
00:04:09.039 --> 00:04:15.520
So, until Vanderbilt or investigators confirm those details, they remain the attacker's claims.
00:04:15.759 --> 00:04:18.399
And that's actually our first sock lesson here.
00:04:18.639 --> 00:04:23.120
During an incident, separate what you know from what you think you know.
00:04:23.360 --> 00:04:29.120
And that might sound obvious, but it's surprisingly difficult when the incident is moving so quickly.
00:04:29.360 --> 00:04:50.959
You'll have endpoint alerts, users calling the help desk, executives asking questions, screenshots circulating online, threat actors posting claims, sometimes journalists calling, social media rumors, someone's cousin who works in another department saying they heard the entire domain was compromised.
00:04:51.120 --> 00:04:54.560
And the SOC has to separate all of that into categories.
00:04:54.800 --> 00:04:59.600
It's either confirmed, likely, possible, unsupported.
00:04:59.839 --> 00:05:01.759
That matters internally too.
00:05:01.920 --> 00:05:10.160
If you're briefing executives, don't say the attacker stole 1 million records, because the ransomware group said they did.
00:05:10.480 --> 00:05:17.040
Say the attackers claimed to have stolen one million records, but we have not yet independently confirmed that.
00:05:17.199 --> 00:05:19.439
And that's a very different statement.
00:05:19.680 --> 00:05:23.199
So let's talk about Vanderbilt's containment decision.
00:05:23.439 --> 00:05:29.439
The university didn't just tell people be careful, they told users to shut systems down.
00:05:29.680 --> 00:05:39.759
That tells us responders were worried enough about potential propagation or continuing attacker activity that disconnecting endpoints was worth the operational disruption.
00:05:39.920 --> 00:05:42.639
And think about what that means at a university.
00:05:42.879 --> 00:05:59.040
Students need laptops for class, faculty need access to research, administrative staff needs business systems, people have grant deadlines, assignments, research data sets, shared drives, cloud storage, scheduling systems.
00:05:59.279 --> 00:06:01.680
Pretty much everything runs through technology.
00:06:01.920 --> 00:06:10.160
By Monday evening, Vanderbilt had restored campus Wi-Fi, but access to most university systems and applications remained restricted.
00:06:10.319 --> 00:06:12.639
Classes were still scheduled to continue.
00:06:12.879 --> 00:06:24.000
By Monday evening, Vanderbilt had restored campus Wi-Fi, but access to most university systems and applications remained restricted, even though classes were still scheduled to continue.
00:06:24.240 --> 00:06:27.680
And that is incident response happening in the real world.
00:06:27.920 --> 00:06:31.040
Security doesn't get to make decisions in a vacuum.
00:06:31.279 --> 00:06:34.399
Every containment action has a business cost.
00:06:34.720 --> 00:06:39.759
But sometimes the potential cost of not containing the incident is much larger.
00:06:40.000 --> 00:06:44.000
And there's something else I personally like about Vanderbilt's instructions.
00:06:44.240 --> 00:06:48.800
They explicitly told users don't troubleshoot the problem yourself.
00:06:48.959 --> 00:06:51.680
And that's good advice during an active incident.
00:06:51.839 --> 00:07:04.240
Because your average employee doesn't know whether the strange pop-up they're looking at is a ransomware note, a malicious script, a legitimate security tool, a phishing prompt, or something completely unrelated.
00:07:04.399 --> 00:07:19.360
And if they start clicking things, opening files, reconnecting network cables, running random commands they found on Google, or trying to restore files themselves, they may destroy useful evidence or allow malicious activity to continue.
00:07:19.600 --> 00:07:25.360
The SOC's job during an incident isn't just technical containment, it's controlling the chaos.
00:07:25.600 --> 00:07:27.600
And that brings us to another lesson.
00:07:27.839 --> 00:07:30.560
You need an out-of-band communication plan.
00:07:30.800 --> 00:07:33.279
Think about the instructions Vanderbilt gave.
00:07:33.519 --> 00:07:41.680
Shut down your university computer, disconnect from systems, then contact the IT service desk using another device or telephone.
00:07:41.920 --> 00:07:57.279
If your organization experiences a major incident, can you still communicate with employees if email is unavailable, Teams is unavailable, Active Directory is unavailable, the
corporate VPN is off, employees' laptops are turned off.
00:07:57.519 --> 00:07:59.839
What is your backup communication plan?
00:08:00.000 --> 00:08:07.439
Is it personal email, SMS, emergency notification platform, phone tree, or a public status page?
00:08:07.680 --> 00:08:18.639
We ask that because in the middle of a ransomware incident, it's not the ideal time to discover that your incident response plan is stored on the same SharePoint site nobody can access.
00:08:18.800 --> 00:08:22.720
And Vanderbilt gives us another interesting example of segmentation.
00:08:22.959 --> 00:08:27.759
They reported that their medical center was not being impacted by the university incident.
00:08:27.920 --> 00:08:30.959
The medical center operates separate IT systems.
00:08:31.199 --> 00:08:43.679
Vanderbilt Health warned its employees not to connect to Vanderbilt University's IT resources and specifically told them not to connect removable media, laptops, or external storage to university systems.
00:08:43.919 --> 00:08:47.759
Patient care continued at Vanderbilt's hospitals and clinics.
00:08:48.000 --> 00:08:50.240
And that's a pretty important distinction.
00:08:50.559 --> 00:09:06.080
Vanderbilt University and Vanderbilt University Medical Center obviously have a close relationship, but from an IT perspective, enough separation exists that an incident affecting the university did not automatically become an incident affecting the hospital.
00:09:06.240 --> 00:09:09.600
And when we're talking about healthcare, that's enormous.
00:09:09.759 --> 00:09:17.600
We've seen ransomware incidents at hospitals where systems go offline, appointments get canceled, and ambulance start getting diverted.
00:09:17.840 --> 00:09:26.720
BYOD itself recently claimed an attack against the University of New Mexico Health Sciences Center that caused major operational disruptions.
00:09:26.960 --> 00:09:31.840
At Vanderbilt, at least so far, the health system has remained operational.
00:09:32.000 --> 00:09:35.120
And that is what segmentation is supposed to accomplish.
00:09:35.360 --> 00:09:40.639
Segmentation doesn't necessarily prevent the original compromise, it just limits the blast radius.
00:09:40.879 --> 00:09:43.200
It's also not just VLANs.
00:09:43.440 --> 00:10:01.600
It's separate identity systems, separate administrative credentials, separate management infrastructure, controlled trust relationships, limited pathways between environments, restrictions on removable media, and procedures telling people do not connect these two environments while we're investigating.
00:10:01.759 --> 00:10:05.600
That's defense in depth working at an organizational level.
00:10:05.840 --> 00:10:10.000
And there's also another really interesting detail in Vanderbilt's warning.
00:10:10.159 --> 00:10:14.799
They specifically told users to watch for changes inside OneDrive and Box.
00:10:15.120 --> 00:10:17.600
That should catch the attention of SOC teams.
00:10:17.759 --> 00:10:22.480
Because ransomware isn't necessarily limited to somebody's local C drive anymore.
00:10:22.720 --> 00:10:28.240
Modern endpoints synchronize enormous amounts of information with cloud storage platforms now.
00:10:28.480 --> 00:10:34.320
If malicious activity modifies thousands of synchronized files, those changes may propagate.
00:10:34.480 --> 00:10:40.000
If credentials or sessions are compromised, the attacker may interact directly with cloud data.
00:10:40.159 --> 00:10:45.919
So when you're responding to ransomware today, don't only ask what happened to this laptop.
00:10:46.240 --> 00:10:50.000
Ask what cloud services was this user connected to?
00:10:50.240 --> 00:10:52.240
What synchronized folders were mounted?
00:10:52.399 --> 00:10:53.759
What tokens were acted?
00:10:54.000 --> 00:10:56.799
Did the attacker modify cloud hosted files?
00:10:57.039 --> 00:10:58.639
Did they create sharing links?
00:10:58.799 --> 00:11:00.240
Did they change permissions?
00:11:00.399 --> 00:11:03.039
Did anything propagate beyond the endpoint?
00:11:03.200 --> 00:11:07.759
Your endpoint and your cloud environment are increasingly part of the same investigation.
00:11:08.000 --> 00:11:10.080
And we also need to talk about backups.
00:11:10.320 --> 00:11:14.240
BYOD claims that it has access to Vanderbilt's backups.
00:11:14.399 --> 00:11:16.720
Again, that is an attacker claim.
00:11:16.879 --> 00:11:18.720
We don't currently know whether that's true.
00:11:18.879 --> 00:11:22.720
But ransomware groups targeting backups is absolutely real.
00:11:22.960 --> 00:11:26.000
Attackers understand the basic ransomware equation.
00:11:26.240 --> 00:11:31.519
If an organization can restore everything quickly, the attacker's leverage disappears.
00:11:31.759 --> 00:11:42.960
So they go after things like backup servers, backup administrators, storage appliances, hypervisors, snapshots, cloud backup accounts, anything that makes recovery possible.
00:11:43.200 --> 00:11:47.600
That's why backup security deserves the same attention as production security.
00:11:47.840 --> 00:11:59.679
Make sure you have different credentials, MFAware supported, network restrictions, immutable storage, offline or logically isolated copies, and regular restoration testing.
00:11:59.840 --> 00:12:04.320
A backup that exists but can't actually be restored isn't a backup at all.
00:12:04.480 --> 00:12:07.440
It's a very expensive emotional support file.
00:12:07.600 --> 00:12:10.320
And here's another interesting thing about this incident.
00:12:10.639 --> 00:12:13.200
BYOD is a new group.
00:12:13.440 --> 00:12:20.960
Researchers quoted by Recorded Futures say they didn't see the group's first victim until around September 29th of this year.
00:12:21.200 --> 00:12:28.639
Now, a little over a month later, they're claiming attacks involving major universities and healthcare organizations.
00:12:29.039 --> 00:12:36.320
Michigan State University recently disclosed suspicious cyber
activity and later appeared on BYOD's leak site.
00:12:36.480 --> 00:12:43.120
The University of New Mexico Health Sciences Center was hit by a disruptive incident that BYOD also claimed.
00:12:43.360 --> 00:12:45.120
And now, Vanderbilt.
00:12:45.279 --> 00:12:50.639
That's another reminder that defenders shouldn't become too fixated on ransomware brand names.
00:12:50.879 --> 00:13:03.919
Groups appear, they disappear, affiliates move, code gets sold, infrastructure gets reused, someone launches a new leak site, the logo changes, but the behavior often doesn't.
00:13:04.080 --> 00:13:14.000
Whether the ransomware note says Lockbit, Akira, Klopp, BYOD, or some group we've never heard of before, the defensive questions are still familiar.
00:13:14.240 --> 00:13:15.279
How did they get in?
00:13:15.519 --> 00:13:17.759
What identity did they compromise?
00:13:18.000 --> 00:13:20.000
What privilege did they obtain?
00:13:20.240 --> 00:13:21.360
Where did they move?
00:13:21.519 --> 00:13:23.279
What data did they access?
00:13:23.440 --> 00:13:25.519
How did they establish persistence?
00:13:25.679 --> 00:13:27.120
What did they exfiltrate?
00:13:27.279 --> 00:13:28.559
What did they encrypt?
00:13:28.720 --> 00:13:30.320
And how do we get them out?
00:13:30.480 --> 00:13:36.799
The name on the ransom note matters for threat intelligence, the behavior matters for incident response.
00:13:37.120 --> 00:13:40.559
So what can SOX learn from Vanderbilt right now?
00:13:40.879 --> 00:13:43.600
First, have the authority to contain.
00:13:43.840 --> 00:13:51.039
If your incident response team believes shutting down hundreds or thousands of systems is necessary, who can authorize that?
00:13:51.360 --> 00:13:52.320
Do you know?
00:13:52.720 --> 00:13:54.799
Can that decision happen quickly?
00:13:54.960 --> 00:14:03.120
Or do you need six executives, legal, communications, and someone's assistant to schedule a teams meeting three hours from now?
00:14:03.360 --> 00:14:07.120
Containment authority should be decided before the incident.
00:14:07.679 --> 00:14:11.120
Second, have an out of band communication plan.
00:14:11.519 --> 00:14:17.919
Know how you're going to talk to employees when corporate communication systems are unavailable or untrusted.
00:14:18.159 --> 00:14:21.360
Third, make security instructions simple.
00:14:21.600 --> 00:14:27.679
Vanderbilt's directions were basically don't click, don't troubleshoot, power off, call us another way.
00:14:27.919 --> 00:14:34.159
During an emergency, that is much better than emailing employees a two-page incident response procedure.
00:14:34.480 --> 00:14:37.039
People need instructions they can remember.
00:14:37.279 --> 00:14:40.879
Fourth, protect your backups and monitor them.
00:14:41.039 --> 00:14:46.399
Your backup infrastructure should be one of the hardest things in your environment for an attacker to reach.
00:14:46.639 --> 00:14:49.919
Fifth, segment critical environments.
00:14:50.159 --> 00:15:00.000
The fact that Vanderbilt's health systems remain separate from the university incident is exactly the outcome you want when one part of a larger organization gets compromised.
00:15:00.320 --> 00:15:04.879
And finally, don't let the threat actor write your incident report.
00:15:05.120 --> 00:15:11.120
Use their claims as intelligence, investigate them, but verify everything independently.
00:15:11.360 --> 00:15:14.080
So here's my challenge for your SOC this week.
00:15:14.399 --> 00:15:19.840
Imagine tomorrow morning you discover ransomware actively spreading through your environment.
00:15:20.080 --> 00:15:23.679
You decide every employee needs to shut down immediately.
00:15:23.919 --> 00:15:26.240
Now, answer these four questions.
00:15:26.559 --> 00:15:29.600
One, who has the authority to make that call?
00:15:29.840 --> 00:15:30.399
2.
00:15:30.639 --> 00:15:32.240
How do you tell everyone?
00:15:32.559 --> 00:15:33.039
3.
00:15:33.360 --> 00:15:36.639
How do they contact the SOC once their computer is offline?
00:15:36.799 --> 00:15:42.000
And four, which parts of your organization continue operating after you pull the plug?
00:15:42.240 --> 00:15:54.799
If those questions are difficult to answer, that's useful information, because you want to figure that out during a tabletop exercise, not while ransomware notes are popping up on employees' laptops.
00:15:55.039 --> 00:15:58.159
And I think that's the biggest lesson from Vanderbilt here.
00:15:58.399 --> 00:16:04.480
Cybersecurity people sometimes focus so heavily on preventing incidents that we forget something.
00:16:04.720 --> 00:16:07.679
Eventually, something is going to get through.
00:16:07.919 --> 00:16:15.440
Maybe it's ransomware, maybe it's a stolen identity, could be a zero day, could be somebody clicks something they shouldn't.
00:16:15.679 --> 00:16:20.320
Your security program isn't judged only by whether the attacker gets in.
00:16:20.480 --> 00:16:24.000
It's also judged by what happens after they get in.
00:16:24.240 --> 00:16:25.679
Can you identify it?
00:16:25.919 --> 00:16:27.120
Can you contain it?
00:16:27.360 --> 00:16:28.639
Can you communicate?
00:16:28.879 --> 00:16:31.519
Can you keep critical operations running?
00:16:31.840 --> 00:16:32.960
Can you restore?
00:16:33.200 --> 00:16:39.200
And can you make decisions quickly enough that a bad incident doesn't become a catastrophic one?
00:16:39.440 --> 00:16:44.000
Right now, Vanderbilt is still working through those questions in real time.
00:16:44.320 --> 00:16:55.440
We don't yet know the complete attack path, we don't know whether all of BYOD's data theft claims are true, and we'll probably learn considerably more over the next several days.
00:16:55.679 --> 00:16:57.360
But one thing is already clear.
00:16:57.600 --> 00:17:02.320
When Vanderbilt believed the threat was serious enough, they told people to shut it down.
00:17:02.559 --> 00:17:08.559
Sometimes the smartest thing a SOC can do isn't another detection rule, it's knowing when to pull the plug.
00:17:08.880 --> 00:17:11.680
And that's a wrap for this episode of the SOC Brief.
00:17:11.759 --> 00:17:21.119
If your team has gone through a major ransomware incident or you've built a good out-of-band communication and containment process, hit us up through our website or on social media.
00:17:21.279 --> 00:17:24.640
And maybe ask your incident response team one question today.
00:17:24.880 --> 00:17:29.759
If we had to shut everything down tomorrow morning, would we know what happens next?
00:17:30.000 --> 00:17:33.839
Keep your eyes open, keep sharpening those skills, and we'll talk soon.
00:17:34.000 --> 00:17:35.440
Stay secure out there.
00:17:35.759 --> 00:17:36.319
Bye.