If you’ve got a robust security system in place and have a team member that has over 15 years of cybersecurity experience, you might feel like you’ve done enough to be safe. Unfortunately, for at least one business, that wasn’t enough. We recently had a chance to chat with Luke Emrich, Director of Incident Response at Tetra Defense. He investigated the above case and shared what he learned as well as some tips for how organizations can better prepare for attacks.
Tetra DefenseTetra helps businesses deal with ransomware attacks, business email compromises, IP theft investigations, and other incidents that need response and investigation. They help businesses improve their cybersecurity posture in accordance with the latest threats (and the latest solutions). As one of Tetra’s case leads, Luke will be one of the first people that a client will speak to when an incident occurs. Together they will work through a game plan and strategy for how to investigate the event, and more importantly, how to recover and how to get the business up and running.
The AttackIt was a Saturday night when it was obvious that something was wrong at a medium-sized SAS company. The 15-year cybersecurity expert we alluded to in the introduction took copious notes from a security engineer late in the afternoon. Server shutdown alerts were coming in one after another and it became clear pretty early on that three separate attacks were happening simultaneously. So this wasn’t one burglar getting into one house, it was a simultaneous attack on three different houses in a neighborhood.
Find the SourceAs Luke and Tetra started looking at the situation, they began with the high-value systems that the threat actor may have used as part of the attack. In this case, the threat actor used a tool called Cobalt Strike.
Cobalt StrikeCobalt Strike is a tool that is often used by the US government and large businesses to emulate the tactics and techniques of a threat actor. In one exercise, one group of “hackers” will use Cobalt Strike to deploy a payload that creates a connection to a server. In the case we’re discussing, the threat actor used Cobalt Strike to deploy ransomware. This allowed the threat actor to go from controlling one computer to five, then ten, and so on exponentially.
Planning and PreparationBecause the company had plans in place for situations like this one, Luke and his team were able to move pretty quickly to try to look at the logs, which the threat actor had been clearing. But because the company had a very granular backup program as part of their planning for attacks, Luke and his team were able to find out when the system was compromised.
Not Just MalwareLuke is quick to point out that ransomware these days is not someone clicking on a bad link that downloads malware. Threat actors these days are likely to have done a lot of reconnaissance and know user names, passwords, and locations of key systems. That’s why it’s important to figure out where the threat actor got in, so you can put protection in place to prevent it from happening again, but then can also restore systems from before that entry. If there aren’t backup protocols in place, restoration won’t be a matter of hours or days, but weeks or months.
Finding the SourceLuke and his team were initially focused on the corporate domain, because that’s where the most important and sensitive data was. But there was a separate domain that had a legacy environment and protocols on it. Even though the threat actor had been clearing the security logs, after a bit of digging Luke and the team were able to see that there had been a Bloodhound detection.
BloodhoundBloodhound is an open source project that was released in 2016. It was originally designed for offense: it looks at your setup to determine how big the attack path problem is for your business. You can think of it as a hammer that you can use to hammer nails to help build doors of security. But hammers can also be used to break down doors, and that’s what the threat actor was doing in this case: using Bloodhound to find weaknesses and the fastest paths to gain what they desired. The use of Bloodhound in this particular case led to the realization of the team that there were local administrative credentials being used (remember the legacy environment we mentioned). Those credentials should have been changed when a new computer system was provisioned, but on the antivirus server that bridged the different domains (legacy vs current) the credentials were not changed. When they followed this path they found that not all the logs at that level had been cleared, which allowed the team to trace back to the original entry point.
Six Days LaterLuke had the initial call with the client on Saturday, but it wasn’t until the following Saturday that Tetra was able to come back with answers as to how the threat actor got in, when they got in, and how the company could use their backups to get back up and running. That meant there were six days of hard teamwork (often around the clock) running down leads and trying to follow clues.
Be PreparedWhile a week seems like a long time, Luke emphasizes that this was actually a short time to restoration because the company had been prepared and followed a predetermined recovery plan. Some of that preparation includes:
- Having a robust security program
- Developing an incident response plan which includes how to act and when
- Incorporating vulnerability management and patching
- Ensuring ongoing monitoring of systems (this could be a security operation center or a product that’s watching for anomalous activity in your systems and potentially implementing containment steps if an event develops)