Blog
/
Network
/
April 5, 2022

How Darktrace Antigena Thwarted Cobalt Strike Attack

Learn how Darktrace's Antigena technology intercepted and delayed a Cobalt Strike intrusion. Discover more cybersecurity news and analyses on Darktrace's blog.
Inside the SOC
Darktrace cyber analysts are world-class experts in threat intelligence, threat hunting and incident response, and provide 24/7 SOC support to thousands of Darktrace customers around the globe. Inside the SOC is exclusively authored by these experts, providing analysis of cyber incidents and threat trends, based on real-world experience in the field.
Written by
Dylan Evans
Default blog image
05
Apr 2022

In December 2021 several CVEs[1] were issued for the Log4j vulnerabilities that sent security teams into a global panic. Threat actors are now continuously scanning external infrastructure for evidence of the vulnerability to deploy crypto-mining malware.[2] However, through December ‘21 – February ‘22, it was ransomware groups that seized the initiative.

Compromise

In January 2022, a Darktrace customer left an external-facing VMware server unpatched allowing Cobalt Strike to be successfully installed. Several IoCs indicate that Cuba Ransomware operators were behind the attack. Thanks to the Darktrace SOC service, the customer was notified of the active threat on their network, and Antigena’s Autonomous Response was able to keep the attackers at bay before encryption events took place.

Initially the VMware server breached two models relating to an anomalous script download and a new user agent both connecting via HTTP. As referenced in an earlier Darktrace blog, both of these models had been seen in previous Log4j exploits. As with all Darktrace models however, the model deck is not designed to detect only one exploit, infection variant, or APT.

Figure 1: Darktrace models breaching due to the malicious script download

Analyst investigation

A PCAP of the downloaded script showed that it contained heavily obfuscated JavaScript. After an OSINT investigation a similar script was uncovered which likely breached the same Yara rules.

Figure 2: PCAP of the Initial HTTP GET request for the Windows Script component

Figure 3: PCAP of the initial HTTP response containing obfuscated JavaScript

Figure 4: A similar script that has been observed installing additional payloads after an initial infection[3]

While not an exact match, this de-obfuscated code shared similarities to those seen when downloading other banking trojans.

Having identified on the Darktrace UI that this was a VMware server, the analyst isolated the incoming external connections to the server shortly prior to the HTTP GET requests and was able to find an IP address associated with Log4j exploit attempts.

Figure 5: Advanced Search logs showing incoming SSL connections from an IP address linked to Log4j exploits

Through Advanced Search the analyst identified spikes shortly prior and immediately after the download. This suggested the files were downloaded and executed by exploiting the Log4j vulnerability.

Antigena response

Figure 6: AI Analyst reveals both the script downloads and the unusual user agent associated with the connections

Figure 7: Antigena blocked all further connections to these endpoints following the downloads

Cobalt Strike

Cobalt Strike is a popular tool for threat actors as it can be used to perform a swathe of MITRE ATT&CK techniques. In this case the threat actor attempted command and control tactics to pivot through the network, however, Antigena responded promptly when the malware attempted to communicate with external infrastructure.

On Wednesday January 26, the DNS beacon attempted to connect to malicious infrastructure. Antigena responded, and a Darktrace SOC analyst issued an alert.

Figure 8: A Darktrace model detected the suspicious DNS requests and Antigena issued a response

The attacker changed their strategy by switching to a different server “bluetechsupply[.]com” and started issuing commands over TLS. Again, Darktrace detected these connections and AI Analyst reported on the incident (Figure 9, below). OSINT sources subsequently indicated that this destination is affiliated with Cobalt Strike and was only registered 14 days prior to this incident.

Figure 9: AI Analyst summary of the suspicious beaconing activity

Simultaneous to these connections, the device scanned multiple internal devices via an ICMP scan and then scanned the domain controller over key TCP ports including 139 and 445 (SMB). This was followed by an attempt to write an executable file to the domain controller. While Antigena intervened in the file write, another Darktrace SOC analyst was issuing an alert due to the escalation in activity.

Figure 10: AI Analyst summary of the .dll file that Antigena intercepted to the Windows/temp directory of the domain controller

Following the latest round of Antigena blocks, the threat actor attempted to change methods again. The VMware server utilised the Remote Access Tool/Trojan NetSupport Manager in an attempt to install further malware.

Figure 11: Darktrace reveals the attacker changing tactics

Despite this escalation, Darktrace yet again blocked the connection.

Perhaps due to an inability to connect to C2 infrastructure, the attack stopped in its tracks for around 12 hours. Thanks to Antigena and the Darktrace SOC team, the security team had been afforded time to remediate and recover from the active threat in their network. Interestingly, Darktrace detected a final attempt at pivoting from the machine, with an unusual PowerShell Win-RM connection to an internal machine. The modern Win-RM protocol typically utilises port 5985 for HTTP connections however pre-Windows 7 machines may use Windows 7 indicating this server was running an old OS.

Figure 12: Darktrace detects unusual PowerShell usage

Cuba Ransomware

While no active encryption appears to have taken place for this customer, a range of IoCs were identified which indicated that the threat actor was the group being tracked as UNC2596, the operators of Cuba Ransomware.[4]

These IoCs include: one of the initially dropped files (komar2.ps1,[5] revealed by AI Analyst in Figure 6), use of the NetSupport RAT,[6] and Cobalt Strike beaconing.[7] These were implemented to maintain persistence and move laterally across the network.

Cuba Ransomware operators prefer to exfiltrate data to their beacon infrastructure rather than using cloud storage providers, however no evidence of upload activity was observed on the customer’s network.

Concluding thoughts

Unpatched, external-facing VMware servers vulnerable to the Log4j exploit are actively being targeted by threat actors with the aim of ransomware detonation. Without using rules or signatures, Darktrace was able to detect all stages of the compromise. While Antigena delayed the attack, forcing the threat actor to change C2 servers constantly, the Darktrace analyst team relayed their findings to the security team who were able to remediate the compromised machines and prevent a final ransomware payload from detonating.

For Darktrace customers who want to find out more about Cobalt Strike, refer here for an exclusive supplement to this blog.

Appendix

Darktrace model detections

Initial Compromise:

  • Device / New User Agent To Internal Server
  • Anomalous Server Activity / New User Agent from Internet Facing System
  • Experimental / Large Number of Suspicious Successful Connections

Breaches from Critical Devices / DC:

  • Device / Large Number of Model Breaches
  • Antigena / Network / External Threat / Antigena File then New Outbound Block
  • Device / SMB Lateral Movement
  • Experimental / Unusual SMB Script Write V2
  • Compliance / High Priority Compliance Model Breach
  • Anomalous Server Activity / Anomalous External Activity from Critical Network Device
  • Experimental / Possible Cobalt Strike Server IP V2

Lateral Movement:

  • Antigena / Network / Insider Threat / Antigena Internal Anomalous File Activity
  • Compliance / SMB Drive Write
  • Anomalous File / Internal / Executable Uploaded to DC
  • Experimental / Large Number of Suspicious Failed Connections
  • Compromise / Suspicious Beaconing Behaviour
  • Antigena / Network / Significant Anomaly / Antigena Breaches Over Time Block
  • Antigena / Network / External Threat / Antigena Suspicious Activity Block
  • Anomalous Connection / High Volume of Connections to Rare Domain
  • Antigena / Network / Significant Anomaly / Antigena Enhanced Monitoring from Server Block

Network Scan Activity:

  • Device / Suspicious SMB Scanning Activity
  • Experimental / Network Scan V2
  • Device / ICMP Address Scan
  • Experimental / Possible SMB Scanning Activity
  • Experimental / Possible SMB Scanning Activity V2
  • Antigena / Network / Insider Threat / Antigena Network Scan Block
  • Device / Network Scan
  • Compromise / DNS / Possible DNS Beacon
  • Device / Internet Facing Device with High Priority Alert
  • Antigena / Network / Significant Anomaly / Antigena Enhanced Monitoring from Server Block

DNS / Cobalt Strike Activity:

  • Experimental / Possible Cobalt Strike Server IP
  • Experimental / Possible Cobalt Strike Server IP V2
  • Antigena / Network / External Threat / Antigena File then New Outbound Block
  • Antigena / Network / External Threat / Antigena Suspicious File Block
  • Anomalous Connection / New User Agent to IP Without Hostname
  • Anomalous File / Script from Rare External Location

MITRE ATT&CK techniques observed

IoCs

Thanks to Brianna Leddy, Sam Lister and Marco Alanis for their contributions.

Footnotes

1.

https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-44228
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-44530
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-45046
https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-4104

2. https://www.toolbox.com/it-security/threat-reports/news/log4j-vulnerabilities-exploitation-attempts

3. https://twitter.com/ItsReallyNick/status/899845845906071553

4. https://www.mandiant.com/resources/unc2596-cuba-ransomware

5. https://www.ic3.gov/Media/News/2021/211203-2.pdf

6. https://threatpost.com/microsoft-exchange-exploited-cuba-ransomware/178665/

7. https://www.bleepingcomputer.com/news/security/microsoft-exchange-servers-hacked-to-deploy-cuba-ransomware/

8. https://gist.github.com/blotus/f87ed46718bfdc634c9081110d243166

Inside the SOC
Darktrace cyber analysts are world-class experts in threat intelligence, threat hunting and incident response, and provide 24/7 SOC support to thousands of Darktrace customers around the globe. Inside the SOC is exclusively authored by these experts, providing analysis of cyber incidents and threat trends, based on real-world experience in the field.
Written by
Dylan Evans

More in this series

No items found.

Blog

/

AI

/

July 6, 2026

NIST Just Proved It: AI Security Can’t Be Solved With Rules

ai security nistDefault blog imageDefault blog image

Static AI guardrails are inherently limited

As organizations adopt generative AI, many still assume that the right set of guardrails will be enough. The problem is you can’t anticipate every way these systems might be misused, abused or attacked. What NIST has done is put a mathematical foundation under that intuition.

In recent research building on Gödel’s incompleteness theorems, which showed that any system built on a fixed set of rules will always have gaps, NIST demonstrates that there is no finite set of guardrails that can be universally robust against adversarial prompts. In plain terms, if your defense is based on a fixed set of rules, there will always be inputs that bypass them. Not because the rules are badly written, but because the problem space is bigger than static rules can ever cover.

This is not new in cybersecurity - detection rules have always had to live with this trade-off. What is different with GenAI is the scale and shape of that problem. These systems are built on human language, and human language is not bounded. It is fluid, contextual and deliberately ambiguous. The number of ways intent can be hidden is effectively limitless. You are not defending against a defined protocol or a fixed exploit chain. You are defending against the entire expressive capacity of people.

So attempting to create a complete set of rules is the wrong starting point. It assumes the problem can be deterministically described. NIST’s work shows that it cannot. Organizations still need a way to manage AI risk, but the traditional approach of defining allowed and disallowed patterns is always going to lag behind what is actually happening. The same input can be benign in one context and risky in another, and static rules struggle to capture that distinction.

The question then is what fills that gap?

AI security must shift from rules to behavior

What's required is a shift in what you are trying to understand. Rules try to describe what should and shouldn't happen. Behavior shows you what is happening. Or to put it another way, if inputs are unbounded and adversaries adapt, the only stable signal is behavior.

In a GenAI context, that means analyzing how an AI model is being used, how prompts evolve over time, how outputs are shaped, and where AI agent interactions start to drift from what is expected. It means moving from static definitions of bad to a more dynamic understanding of intent.

Instead of trying to predict every bad prompt, you focus on identifying when behavior starts to move outside expected norms. Instead of asking whether a single input matches a rule, you ask whether the overall pattern of activity makes sense for the system and how it’s being used.

Guardrails remain important but they are only one layer

This does not eliminate the need for guardrails. They still play a role. But they will never address the entire problem space and are simply one part of your defense in depth approach.

NIST’s proof is useful because it makes this explicit. It removes the assumption that with enough effort, a complete rule set is achievable. It isn’t.

Once you accept that, the shift becomes unavoidable. This is no longer a problem of writing better rules, but of understanding behavior in a space where the possible inputs are effectively unbounded.

For security leaders, that changes the nature of the problem. It is less about defining what should be allowed, and more about recognizing when something is no longer consistent with expected behavior.

That does not remove the need for guardrails, but it does change their role. They set boundaries, but they do not define understanding. The gap between the two is where risk now sits.

In the end, this is what “can’t be solved with rules” really means. Rules will always leave gaps, and those gaps are not theoretical. They show up in how systems actually behave Not what we expect them to do, or what we intended them to do, but what they are doing in practice. That is where the signal is, and increasingly, that is where the security problem sits.

References:

https://www.nist.gov/news-events/news/2026/06/nist-mathematical-proof-supports-transition-continuous-monitor-and-update

https://ieeexplore.ieee.org/document/11475847

Continue reading
About the author
Andrew Hollister
Principal Solutions Engineer, Cyber Technician

Blog

/

AI

/

July 1, 2026

5 Ways AI is changing traditional security models according to modern CISOs

Default blog imageDefault blog image

The Reality of Securing AI in Motion

Traditional security tools were built for environments defined by fixed rules and predictable workflows. But AI behavior is non-deterministic. The same prompt can produce different outcomes, and risk often emerges gradually as AI behavior adapts, and permissions drift over time. This creates a constantly shifting environment where security teams are working to define control in a system that resists stability. “In AI security, yesterday's priorities can become tomorrow's blind spots. The landscape shifts that fast,” warned the SVP and Head of Technology and Cybersecurity of a real estate investment trust. Conventional approaches, which rely on establishing and maintaining a steady baseline, struggle to keep up with that level of change.

At the same time, AI adoption is accelerating across organizations, often faster than security teams can implement the controls needed to manage it. “The car is being built while it’s already on the road,” explained the CISO of a global private fund administrator. “The threats we're securing against today won't be the threats we're facing tomorrow. What kept us up three months ago looks nothing like what we're dealing with today.”

As businesses move quickly to unlock value from AI, security teams are left closing gaps in real time, while also facing adversaries who are using AI to make their attacks more scalable, adaptive, and difficult to detect. In this recent roundtable discussion of CISOs and security leaders, five themes emerged around AI cyber risk.  

1. AI agents with human access but no human judgment

In Darktrace’s 2026 State of AI Cybersecurity report, 96% of the surveyed security professionals agree that AI significantly improves the speed and efficiency with which they work. Yet, 92% admitted that they’re concerned with the security implications of the use of AI agents across their workforce.

AI agents now operate with human-level permissions across systems, acting at machine speed, orchestrating actions across platforms, and making decisions without the judgment or caution a person would apply. Unlike human users, they cannot be expected to pause and question whether a given action is appropriate.

Their identities are also difficult to inventory, govern, and audit. As agents become easier to deploy than legacy IT systems ever were, organizations are quickly losing track of what is running, what it has access to, and what it is doing. This creates a growing class of highly privileged, autonomous actors operating without the visibility or oversight that traditional identity and access controls were designed to provide.“While AI adoption is critical to running a modern business, AI alone can’t solve all our cybersecurity challenges,” said a global financial sector CISO. “We still need think critically and use human judgement. Those are two things AI can’t do.”

This lack of human judgment becomes especially risky as new architectures, such as Model Context Protocol (MCP), can expand how agents connect to data, tools, and external systems. By design, MCP enables agents to dynamically discover and interact with new resources, increasing flexibility but also introducing new pathways for unintended access, data exposure, or abuse if not properly governed.

The CISO of a fund administrator highlighted one emerging vector as an example: rogue MCP servers. “Our developers want to move quickly and bring value to the business, but technologies like these can unintentionally expose sensitive data in ways that would never have happened before.”

2. Increased digital complexity and expanded attack surface

AI activity rarely stays contained. A single prompt can trigger a chain of actions across networks, email, cloud infrastructure, SaaS platforms, endpoints, identity systems, and development environments, spanning systems that were never designed to be secured as a single, connected flow. This expands both the scale and complexity of what security teams need to monitor and defend.

Yet no single control has visibility across that entire chain. “You can’t defend effectively what you can’t see,” cautioned the private fund administrator CISO. As AI-driven activity moves fluidly across environments, gaps in coverage become inevitable, creating blind spots that attackers can exploit.

Threat actors are already capitalizing on this lack of visibility. “Threat actors have advanced their use of generative AI to launch more convincing phishing campaigns, automate social engineering, and scale attacks with greater precision down to the individual level,” said the SVP of Technology and Cybersecurity for the real estate investment trust. What was once manual and targeted can now be automated and personalized at scale, making attacks harder to detect and easier to execute.

At the same time, the pace of exploitation is accelerating. As a global CISO operating across 40+ countries described it: “Zero-day vulnerabilities are no longer zero day; it’s minus one day. By the time you get to it and address it, it’s already a problem.” By the time risk is identified, it has often already been realized.

The result is a rapidly expanding and increasingly interconnected attack surface that challenges security teams to maintain visibility, context, and control across AI-driven activity.

3. Shadow AI is already everywhere

76% of organizations now cite shadow AI as a problem, one that is spreading through organizations in ways that are hard to track and even harder to control.

Employees are experimenting with publicly available Gen AI tools. Teams are spinning up low-code automations on their own. SaaS providers are quietly embedding AI into existing products. Developers are plugging AI services directly into workflows, often without pausing to consider what that exposure means.

The result is a lack of visibility into:

  • What AI tools are being used
  • What data those tools can access
  • Where prompts and outputs are going
  • Which AI agents are interacting with enterprise systems

The SVP of Cybersecurity at a real estate investment trust described the shift: “Before, I was worried about someone sending data erroneously to their personal email. Now we have all these agents online that people are utilizing, and we’re looking at those vectors as well.” For security teams, this means operating without a complete view of how AI is being used, what it can access, and where risk may already be emerging.

4. Built-in guardrails are not enough

Organizations often assume that native AI guardrails or provider-level controls are sufficient to manage AI risk. But securing AI requires ongoing visibility, oversight, and governance, not just controls configured at deployment. "It’s a misconception that adopting AI is going to solve all your problems,” warns a global financial services CISO.

Security leaders are increasingly recognizing the limitations of these controls as:

  • Fragmented and difficult to enforce consistently across multiple AI systems, workflows, and environments
  • Ambiguous in terms of accountability due to shared responsibility for AI governance between IT, security, developers, business teams, and third-party providers
  • Limited in end-to-end oversight, leaving gaps that stretch from the initial prompt all the way through to the downstream impact of an agent's actions

Securing AI demands more than simple prompt filtering or static policy enforcement. It requires understanding intent, behavior, and context across both human and AI activity.

The next phase of cybersecurity: securing AI

To safely and responsibly adopt AI at scale, organizations need a new operational model for cybersecurity that’s capable of:

• Understanding AI behavior

• Identifying risk in real time

• Maintaining governance without slowing innovation

The CSO of a $10 billion municipal utility organization described the challenge with precision: “We have to move at the speed of innovation and risk, because both are accelerating faster than ever.”

Embrace AI with confidence with Darktrace / SECURE AI

Darktrace has introduced Darktrace / SECURE AI™, a new product within the Darktrace ActiveAI Security Platform™  ,designed to provide enterprise-wide security for AI by applying industry leading behavioral analysis to how prompts, agents, and AI systems are used.

Darktrace / SECURE AITM delivers real-time visibility and control across Enterprise and SaaS GenAI prompts, AI agent identities, development and production environments, and Shadow AI - detecting even subtle misuse, misconfiguration, and drift that traditional, rule-based controls simply do not understand. By interpreting context and intent across humans and machines, Darktrace enables organizations to adopt AI at scale without introducing unmanaged risk

What makes this possible is Darktrace’s decade-long maturity and expertise in behavioral understanding and AI-native cybersecurity. Achieved with Self-Learning AI that has been proven across more than 10,000 organizations, Darktrace understands what “normal” looks like for a business, across its users, systems, and now AI, so that meaningful deviations can be detected and acted on before they become incidents.

With one CISO describing Darktrace’s Self-Learning AI as “a leap forward compared to other tools” and another as a “force multiplier,” the technology can interpret ambiguous interactions, understand how access accumulates over time, and recognize when behavior, human or machine, begins to drift.

“Strategically, we’re looking to gain more visibility into how AI is operating across the environment and achieve greater control over what AI should be allowed to access and do,” shared the CISO at a private fund administrator.  

“What I’ve seen from Darktrace / SECURE AI is extremely promising. I have tremendous confidence in Darktrace’s vision for where this is headed and its ability to execute on this new solution.”

Continue reading
About the author
The Darktrace Community
Your data. Our AI.
Elevate your network security with Darktrace AI