August 22, 2022

Emotet Resurgence: Cross-Industry Analysis

Technical insights on the Emotet resurgence in 2022 across various client environments, industries, and regions.

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

Eugene Chua

Cyber Security Analyst

Aug 2022

Introduction

Last year provided further evidence that the cyber threat landscape remains both complex and challenging to predict. Between uncertain attribution, novel exploits and rapid malware developments, it is becoming harder to know where to focus security efforts. One of the largest surprises of 2021 was the re-emergence of the infamous Emotet botnet. This is an example of a campaign that ignored industry verticals or regions and seemingly targeted companies indiscriminately. Only 10 months after the Emotet takedown by law enforcement agencies in January, new Emotet activities in November were discovered by security researchers. These continued into the first quarter of 2022, a period which this blog will explore through findings from the Darktrace Threat Intel Unit.

Dating back to 2019, Emotet was known to deliver Trickbot payloads which ultimately deployed Ryuk ransomware strains on compromised devices. This interconnectivity highlighted the hydra-like nature of threat groups wherein eliminating one (even with full-scale law enforcement intervention) would not rule them out as a threat nor indicate that the threat landscape would be any more secure.

When Emotet resurged, as expected, one of the initial infection vectors involved leveraging existing Trickbot infrastructure. However, unlike the original attacks, it featured a brand new phishing campaign.

Figure 1: Distribution of observed Emotet activities across Darktrace deployments

Although similar to the original Emotet infections, the new wave of infections has been classified into two categories: Epochs 4 and 5. These had several key differences compared to Epochs 1 to 3. Within Darktrace’s global deployments, Emotet compromises associated to Epoch 4 appeared to be the most prevalent. Affected customer environments were seen within a large range of countries (Figure 1) and industry verticals such as manufacturing and supply chain, hospitality and travel, public administration, technology and telecoms and healthcare. Company demographics and size did not appear to be a targeting factor as affected customers had varying employee counts ranging from less than 250, to over 5000.

Key differences between Epochs 1-3 vs 4-5

Based on wider security research into the innerworkings of the Emotet exploits, several key differences were identified between Epochs 4/5 and its predecessors. The newer epochs used:

· A different Microsoft document format (OLE vs XML-based).

· A different encryption algorithm for communication. The new epochs used Elliptic Curve Cryptograph (ECC) [1] with public encryption keys contained in the C2 configuration file [2]. This was different from the previous Rivest-Shamir-Adleman (RSA) key encryption method.

· Control Flow Flattening was used as an obfuscation technique to make detection and reverse engineering more difficult. This is done by hiding a program’s control flow [3].

· New C2 infrastructure was observed as C2 communications were directed to over 230 unique IPs all associated to the new Epochs 4 and 5.

In addition to the new Epoch 4 and 5 features, Darktrace detected unsurprising similarities in those deployments affected by the renewed campaign. This included self-signed SSL connections to Emotet’s new infrastructure as well as malware spam activities to multiple rare external endpoints. Preceding these outbound communications, devices across multiple deployments were detected downloading Emotet-associated payloads (algorithmically generated DLL files).

Emotet Resurgence Campaign

Figure 2: Darktrace’s Detection Timeline for Emotet Epoch 4 and 5 compromises

1. Initial Compromise

The initial point of entry for the resurgence activity was almost certainly via Trickbot infrastructure or a successful phishing attack (Figure 2). Following the initial intrusion, the malware strain begins to download payloads via macro-ladened files which are used to spawn PowerShell for subsequent malware downloads.

Following the downloads, malicious communication with Emotet’s C2 infrastructure was observed alongside activities from the spam module. Within Darktrace, key techniques were observed and documented below.

2. Establish Foothold: Binary Dynamic-link library (.dll) with algorithmically generated filenames

Emotet payloads are polymorphic and contain algorithmically generated filenames . Within deployments, HTTP GET requests involving a suspicious hostname, www[.]arkpp[.]com, and Emotet related samples such as those seen below were observed:

· hpixQfCoJb0fS1.dll (SHA256 hash: 859a41b911688b00e104e9c474fc7aaf7b1f2d6e885e8d7fbf11347bc2e21eaa)

· M0uZ6kd8hnzVUt2BNbRzRFjRoz08WFYfPj2.dll (SHA256 hash: 9fbd590cf65cbfb2b842d46d82e886e3acb5bfecfdb82afc22a5f95bda7dd804)

· TpipJHHy7P.dll (SHA256 hash: 40060259d583b8cf83336bc50cc7a7d9e0a4de22b9a04e62ddc6ca5dedd6754b)

These DLL files likely represent the distribution of Emotet loaders which depends on windows processes such as rundll32[.]exe and regsvr32[.]exe to execute.

3. Establish Foothold: Outbound SSL connections to Emotet C2 servers

A clear network indicator of compromise for Emotet’s C2 communication involved self-signed SSL using certificate issuers and subjects which matched ‘CN=example[.]com,OU=IT Department,O=Global Security,L=London,ST=London,C=GB’ , and a common JA3 client fingerprint (72a589da586844d7f0818ce684948eea). The primary C2 communications were seen involving infrastructures classified as Epoch 4 rather than 5. Despite encryption in the communication content, network contextual connection details were sufficient for the detection of the C2 activities (Figure 3).

Figure 3: UI Model Breach logs on download and outbound SSL activities.

Outbound SSL and SMTP connections on TCP ports 25, 465, 587

An anomalous user agent such as, ‘Microsoft Outlook 15.0’, was observed being used for SMTP connections with some subject lines of the outbound emails containing Base64-encoded strings. In addition, this JA3 client fingerprint (37cdab6ff1bd1c195bacb776c5213bf2) was commonly seen from the SSL connections. Based on the set of malware spam hostnames observed across at least 10 deployments, the majority of the TLDs were .jp, .com, .net, .mx, with the Japanese TLD being the most common (Figure 4).

Figure 4: Malware Spam TLDs observed in outbound SSL and SMTP

Plaintext spam content generated from the spam module were seen in PCAPs (Figure 5). Examples of clear phishing or spam indicators included 1) mismatched personal header and email headers, 2) unusual reply chain and recipient references in the subject line, and 3) suspicious compressed file attachments, e.g. Electronic form[.]zip.

Figure 5: Example of PCAP associated to SPAM Module

4. Accomplish Mission

The Emotet resurgence also showed through secondary compromises involving anomalous SMB drive writes related to CobaltStrike. This consistently included the following JA3 hash (72a589da586844d7f0818ce684948eea) seen in SSL activities as well as SMB writes involving the svchost.exe file.

Darktrace Detection

The key DETECT models used to identify Emotet Resurgence activities were focused on determining possible C2. These included:

· Suspicious SSL Activity

· Suspicious Self-Signed SSL

· Rare External SSL Self-Signed

· Possible Outbound Spam

File-focused models were also beneficial and included:

· Zip or Gzip from Rare External Location

· EXE from Rare External Location

Darktrace’s detection capabilities can also be shown through a sample of case studies identified during the Threat Research team’s investigations.

Case Studies

Darktrace’s detection of Emotet activities was not limited by industry verticals or company sizing. Although there were many similar features seen across the new epoch, each incident displayed varying techniques from the campaign. This is shown in two client environments below:

When investigating a large customer environment within the public administration sector, 16 different devices were detected making 52,536 SSL connections with the example[.]com issuer. Devices associated with this issuer were mainly seen breaching the same Self-Signed and Spam DETECT models. Although anomalous incoming octet-streams were observed prior to this SSL, there was no clear relation between the downloads and the Emotet C2 connections. Despite the total affected devices occupying only a small portion of the total network, Darktrace analysts were able to filter against the much larger network ‘noise’ and locate detailed evidence of compromise to notify the customer.

Darktrace also identified new Emotet activities in much smaller customer environments. Looking at a company in the healthcare and pharmaceutical sector, from mid-March 2022 a single internal device was detected making an HTTP GET request to the host arkpp[.]com involving the algorithmically-generated DLL, TpipJHHy7P.dll with the SHA256 hash: 40060259d583b8cf83336bc50cc7a7d9e0a4de22b9a04e62ddc6ca5dedd6754b (Figure 6).

Figure 6: A screenshot from VirusTotal, showing that the SHA256 hash has been flagged as malicious by other security vendors.

After the sample was downloaded, the device contacted a large number of endpoints that had never been contacted by devices on the network. The endpoints were contacted over ports 443, 8080, and 7080 involving Emotet related IOCs and the same SSL certificate mentioned previously. Malware spam activities were also observed during a similar timeframe.

The Emotet case studies above demonstrate how autonomous detection of an anomalous sequence of activities - without depending on conventional rules and signatures - can reveal significant threat activities. Though possible staged payloads were only seen in a proportion of the affected environments, the following outbound C2 and malware spam activities involving many endpoints and ports were sufficient for the detection of Emotet.

If present, in both instances Darktrace’s Autonomous Response technology, RESPOND, would recommend or implement surgical actions to precisely target activities associated with the staged payload downloads, outgoing C2 communications, and malware spam activities. Additionally, restriction to the devices’ normal pattern of life will prevent simultaneously occurring malicious activities while enabling the continuity of normal business operations.

Conclusion

· The technical differences between past and present Emotet strains emphasizes the versatility of malicious threat actors and the need for a security solution that is not reliant on signatures.

· Darktrace’s visibility and unique behavioral detection continues to provide visibility to network activities related to the novel Emotet strain without reliance on rules and signatures. Key examples include the C2 connections to new Emotet infrastructure.

· Looking ahead, detection of C2 establishment using suspicious DLLs will prevent further propagation of the Emotet strains across networks.

· Darktrace’s AI detection and response will outpace conventional post compromise research involving the analysis of Emotet strains through static and dynamic code analysis, followed by the implementation of rules and signatures.

Thanks to Paul Jennings and Hanah Darley for their contributions to this blog.

Appendices

Model breaches

· Anomalous Connection / Anomalous SSL without SNI to New External

· Anomalous Connection / Application Protocol on Uncommon Port

· Anomalous Connection / Multiple Connections to New External TCP Port

· Anomalous Connection / Multiple Failed Connections to Rare Endpoint

· Anomalous Connection / Multiple HTTP POSTs to Rare Hostname

· Anomalous Connection / Possible Outbound Spam

· Anomalous Connection / Rare External SSL Self-Signed

· Anomalous Connection / Repeated Rare External SSL Self-Signed

· Anomalous Connection / Suspicious Expired SSL

· Anomalous Connection / Suspicious Self-Signed SSL

· Anomalous File / Anomalous Octet Stream (No User Agent)

· Anomalous File / Zip or Gzip from Rare External Location

· Anomalous File / EXE from Rare External Location

· Compromise / Agent Beacon to New Endpoint

· Compromise / Beacon to Young Endpoint

· Compromise / Beaconing Activity To External Rare

· Compromise / New or Repeated to Unusual SSL Port

· Compromise / Repeating Connections Over 4 Days

· Compromise / Slow Beaconing Activity To External Rare

· Compromise / SSL Beaconing to Rare Destination

· Compromise / Suspicious Beaconing Behaviour

· Compromise / Suspicious Spam Activity

· Compromise / Suspicious SSL Activity

· Compromise / Sustained SSL or HTTP Increase

· Device / Initial Breach Chain Compromise

· Device / Large Number of Connections to New Endpoints

· Device / Long Agent Connection to New Endpoint

· Device / New User Agent

· Device / New User Agent and New IP

· Device / SMB Session Bruteforce

· Device / Suspicious Domain

· Device / Suspicious SMB Scanning Activity

‍

For Darktrace customers who want to know more about using Darktrace to triage Emotet, refer here for an exclusive supplement to this blog.

References

[1] https://blog.lumen.com/emotet-redux/

[2] https://blogs.vmware.com/security/2022/03/emotet-c2-configuration-extraction-and-analysis.html

[3] https://news.sophos.com/en-us/2022/05/04/attacking-emotets-control-flow-flattening/

Written by

Eugene Chua

Cyber Security Analyst

Inside the SOC

Written by

Eugene Chua

Cyber Security Analyst

Specialist Security Researcher

Watch the NIS2 Webinar

Blog

Network

September 25, 2026

A Chain Reaction: Blockchain-Hosted Infostealer Campaign Targets Windows and macOS

Key Insights

Darktrace detected a blockchain-hosted infostealer campaign targeting Windows and macOS devices across multiple customer environments.
The campaign combined ClickFix social engineering with trusted services and decentralized blockchain infrastructure to support malware delivery and C2 activity.
Compromised devices were observed connecting to rare and unusual external endpoints, including DGA C2 domains, blockchain-related endpoints, and cryptocurrency mining infrastructure.
The activity was associated with information-stealing malware strains including Atomic macOS Stealer (AMOS), Lumma, Rhadamanthys, Vidar, and Phexia.
Darktrace identified anomalous device behavior, beaconing patterns, rare external connections, cryptomining activity, and suspicious TLS/SSL communications without relying solely on prior knowledge or static indicators of compromise.
The campaign highlights how attackers are increasingly using legitimate and decentralized infrastructure to make detection, disruption, and attribution more challenging for defenders.

The Infostealer Ecosystem

The information stealer malware ecosystem continues to grow in value for threat actors across the digital threat landscape. Infostealers are increasingly delivered through Malware-as-a-Service (MaaS) operating models, distributed through affiliate networks, and designed to withstand infrastructure takedowns. This resilience was demonstrated by the recent takedown of Lumma Stealer malicious domains by Microsoft’s Digital Crimes Unit (DCU) [1].

Infostealers are used to gather and exfiltrate sensitive information, including non-human identity (NHI) data, from compromised systems across cloud, Software-as-a-Service (SaaS), Virtual Private Network (VPN), and development environments. They can also support ransomware operations by expanding the credentials and access paths available to threat actors, contributing to the high volume of identity-based attacks observed across the broader threat landscape [2][3].

Darktrace’s Observations of ClickFix and Infostealers

Throughout 2026, Darktrace has observed multiple campaigns using ClickFix social engineering to trick users into carrying out malicious actions and downloading initial payloads, including information stealers. More recently, Darktrace’s Threat Research team identified a specific ClickFix campaign involving a blockchain-hosted infostealer targeting Windows and macOS devices.

Darktrace identified affected customer environments across Europe, the United States, Asia, and the Middle East where blockchain-hosted infostealer malware appears to have been delivered to compromised systems following likely ClickFix-driven initial access. Darktrace investigated the activity and found that decentralized blockchain infrastructure, alongside widely trusted legitimate services, was used to support malware delivery and information theft across Windows and macOS systems.

Following initial access, compromised systems established C2 communication, with C2 configuration and payloads hosted on public blockchain infrastructure. The ultimate objective appears to be credential and cryptocurrency theft through the deployment of information stealers such as Atomic macOS Stealer (AMOS), Lumma, Rhadamanthys, and Vidar [5][6][7].

Darktrace’s Investigation

Affected devices across the Darktrace customer base were observed making outbound connections to rare external endpoints in patterns consistent with beaconing and C2 activity. Darktrace primarily detected devices making repeated connections to algorithmically generated domains (DGA) such as hf98x4d[.]site [8]. In many cases, these domains were linked through open-source intelligence (OSINT) to information-stealing malware families including AMOS and Phexia [5][6][7][8][9].

In multiple cases, devices were also observed connecting to blockchain-related endpoints, such as polygon[.]drpc[.]org, as well as legitimate public services, including GitHub. The use of decentralized blockchain infrastructure and trusted services such as GitHub to facilitate malware distribution and C2 activity can make disruption and attribution significantly more difficult for defenders.

Darktrace alsodetected a significant proportion of impacted devices making outboundconnections to cryptocurrency mining infrastructure associated with thelegitimate open-source XMRig mining software and the HashVault mining pool,including pool.hashvault[.]pro and donate[.]ssl[.]xmrig[.]com, which wereabused by the attackers, indicating, includingpool.hashvault[.]pro and donate[.]ssl[.]xmrig[.]com, indicating active cryptominingon compromised systems.

In one case, mining activity was observed before and during connections to the DGA endpoint hf98x4d[.]site. Due to its highly anomalous nature, Darktrace's Real-Time AI Analyst autonomously investigated the activity as it occurred, correlating the two events into a single cryptocurrency mining incident and providing comprehensive visibility into the broader attack.

‍

Figure 1: Real-Time AI Analyst investigation of suspicious SSL and C2 communications with hf98x4d[.]site over port 443.

‍

Figure 2: Real-Time AI Analyst investigation into cryptocurrency mining activity involving pool[.]hashvault[.]pro over SSL on port 443.

‍

Around the same time, Darktrace identified the same device initiating connections to the GitHub endpoint release-assets[.]githubusercontent[.]com while continuing to make repeated connections to hf98x4d[.]site.

‍

Figure 3: Darktrace's detection of an affected device connecting to a GitHub endpoint between repeated connections to the anomalous external endpoint hf98x4d[.]site.

On the network of another customer, Darktrace observed an affected device making highly unusual outbound connections consistent with beaconing activity. The device initiated multiple connections over port 443 to the external hostname polygon[.]drpc[.]org. According to OSINT, this hostname is a Remote Procedure Call (RPC) endpoint provided by dRPC, a legitimate service enabling decentralized applications (dApps), cryptocurrency wallets, and developer tools to interact with the Polygon blockchain [10].

The same device was later observed making repeated TLS/SSL connections to the previously mentioned DGA C2 domain. In addition, it made outbound connections to the external IP 195.242.214[.]34 over destination port 51820, an endpoint associated with the ProtonVPN service. Collectively, these connections to blockchain-related infrastructure, the DGA C2 domain, and ProtonVPN-associated infrastructure suggested the device had been affected by the campaign.

Conclusion

This campaign demonstrates how attackers can combine ClickFix social engineering with trusted services and decentralized blockchain infrastructure to create a resilient, cross-platform malware delivery chain. By using services such as GitHub alongside blockchain RPC endpoints and rapidly replaceable DGA domains, the activity can blend into legitimate traffic while making infrastructure disruption and attribution more difficult.

For defenders, it’s a reminder that trusted infrastructure does not automatically mean trusted activity. Security teams should look for the behaviors surrounding these connections, including unusual outbound communication, repeated beaconing, unexpected access to blockchain services, suspicious TLS/SSL activity and cryptomining. In this campaign, Darktrace identified and correlated these deviations without depending solely on previously known indicators, providing visibility as affected devices moved between legitimate services, decentralized infrastructure and malicious C2 endpoints

Credit to Nahisha Nobregas (Associate Principal Cyber Analyst), Manoel Kadja (Senior Cyber Analyst)

Edited by Ryan Traill (Content Manager)

Appendices

Darktrace Model Detections

▪ Compromise / Beaconing Activity To External Rare

▪ Compromise / Beacon to Young Endpoint

▪ Compromise / Fast Beaconing to DGA

▪ Compromise / High Volume of Connections with Beacon Score

▪ Compromise / DGA Beacon

▪ Compromise / Slow Beaconing Activity To External Rare

▪ Compromise / Agent Beacon (Long Period)

▪ Compromise / Agent Beacon (Medium Period)

▪ Compromise / Sustained SSL or HTTP Increase

▪ Compromise / Large Number of Suspicious Failed Connections

▪ Compromise / SSL Beaconing to Rare Destination

▪ Compromise / Beacon for 4 Days

▪ Compromise / High Priority Crypto Currency Mining

▪ Compromise / Monero Mining

▪ Device / Long Agent Connection to New Endpoint

▪ Device / New Connections On Suspicious Port

▪ Anomalous Connection / High Volume of Connections to Rare Domain

‍

List of Indicators of Compromise (IoCs)

Indicator	Description
hf98x4d[.]site	C2 Endpoint (Hostname)
sj98xe4[.]xyz	C2 Endpoint (Hostname)
citcix6[.]xyz	C2 Endpoint (Hostname)
bduwih8[.]pro	C2 Endpoint (Hostname)

‍

MITRE ATT&CK Mapping

Tactic (ID)	Technique
Persistence (T1176)	Browser Extensions (T1176.001)
Persistence (T1176)	Software Extensions
Command and Control (T1071)	Web Protocols (T1071.001)
Command and Control (T1568)	Domain Generation Algorithms (T1568.002)
Command and Control (T1071)	Application Layer Protocol
Command and Control (T1102)	One-Way Communication (T1102.003)
Command and Control (T1571)	Non-Standard Port
Command and Control (T1104)	Multi-Stage Channels
Command and Control (T1573)	Encrypted Channel
Command and Control (T1008)	Fallback Channels
Initial Access ICS (T0862)	Supply Chain Compromise
Command and Control ICS (T0885)	Commonly Used Port
Collection (T1185)	Browser Session Hijacking
Impact (T1496)	Compute Hijacking (T1496.001)
Impact (T1496)	Resource Hijacking
Command and Control (T1071)	Publish/Subscribe Protocols (T1071.001)
Lateral Movement (T1210)	Exploitation of Remote Services

‍

References:

1. https://www.microsoft.com/en-us/security/blog/2025/05/21/lumma-stealer-breaking-down-the-delivery-techniques-and-capabilities-of-a-prolific-infostealer/

2. https://spycloud.com/resource/report/spycloud-annual-identity-exposure-report-2026/

3. https://www.darktrace.com/blog/why-trust-is-the-new-attack-surface-darktraces-mid-year-threat-update-2026

4. https://www.darktrace.com/blog/unpacking-clickfix-darktraces-detection-of-a-prolific-social-engineering-tactic

5. https://abekweng.medium.com/inside-a-blockchain-hosted-malware-campaign-targeting-windows-and-macos-f5bcdeffed66

6. https://cloud.google.com/blog/topics/threat-intelligence/unc5142-etherhiding-distribute-malware

7. https://haveibeensquatted.com/blog/from-typosquatting-to-macos-backdoor-clickfix-blockchain-c2

8. https://www.virustotal.com/gui/domain/hf98x4d.site/community

9. https://x.com/FABO97662188/status/2074125545026244795

10. https://www.virustotal.com/gui/url/b0e5c51a411065864119c305fddf218b7c120731f655932cc1c3307ad5b43f94/gti-summary

About the author

Nahisha Nobregas

SOC Analyst

Blog

September 24, 2026

Detecting Rogue Agent Behavior in the Enterprise

Agents cannot be trusted to perform tasks in the way we intend them to. They may cheat to accomplish their objective, and they may employ hacking methods along the way. Researchers from Darktrace Signal Labs induced cheating behavior from agents deployed in a test environment to analyze the agents’ activities and to assess the performance of the Darktrace platform. Agents frequently resorted to hacking to cheat on their assigned task. The visibility and behavioral profiling provided by both Darktrace / SECURE AI and Darktrace / HYBRID NETWORK ensured extensive detection coverage of the agents’ misaligned activities.

Key Takeaways:

Darktrace Researchers deployed agents in a simulated corporate environment and asked them to solve an impossible challenge. The agents independently turned to traditional hacking techniques to reach their objective. No one instructed them to do this, and no attacker was involved.
Continuously monitoring behavior against a baseline of what is normal for each organization is critical to build trust in enterprise AI.
If an agent may resort to intrusion techniques simply because its assigned task is not possible, then every organization deploying agents within real business processes is at risk. Darktrace / SECURE AI and Darktrace / HYBRID NETWORK identified the agents’ misaligned behavior in real time, with Autonomous Response disrupting it at an early stage.

Introduction: Understanding the Threat of Hacking by Agents

Over the last few months, there has been a surge in reporting [1, 2, 3, 4, 5, 6, 7, 8, 9] of LLM-powered agents engaging in unauthorized hacking activity during evaluations of their capabilities. In several of these cases, including the OpenAI / Hugging Face incident [10], agents engaged in hacking activity as a means of cheating on their evaluations.

To better understand the threat of unauthorized hacking by agents, and the role of Darktrace in combatting it, researchers from Darktrace Signal Labs deployed agents powered by frontier models, including OpenAI’s Daybreak Red models, in simulated, corporate networks. Cheating behavior was evoked through the inclusion of impossible tasks in a coding challenge.

Regardless of the underlying model, agents employed hacking methods to ensure an optimal outcome on the challenge. Darktrace / HYBRID NETWORK and Darktrace / SECURE AI identified the agents’ deviant activities, with inhibitive actions being autonomously taken in the early stages to disrupt the agents’ progression.

Setting the Stage

As part of the research, a Pi agent harness [11] was deployed on a Linux server in Darktrace’s testing environment, which simulates a corporate Active Directory (AD) environment. The same environment included a benchmark server hosting the coding exercise’s contents and grader, as well as various other servers, such as a domain controller and a build server. The model powering the Pi agent varied across tests, with Anthropic’s Opus 4.6 model and OpenAI’s GPT‑5.5‑Cyber model being most widely used.

The Pi agent running on the Linux server was assigned the task of achieving full marks on a coding exercise. The activities of the agent were monitored closely in real-time to minimize the risk of escape from the environment. Monitoring of agent activity was achieved through monitoring of prompts, tool calls, network telemetry, and process telemetry.

Both Darktrace / HYBRID NETWORK and Darktrace / SECURE AI were deployed in the testing environment, with HYBRID NETWORK providing real-time coverage of process creation activity and network traffic, and SECURE AI providing real-time coverage of agent sessions and their contents.

Inducing Cheating Behavior

Cheating behavior, broadly understood as task completion in a way which “perverts the spirit of the designer’s intent” [12], is increasingly being reported as a common behavior of frontier models during evaluations of their capabilities [13].

Misaligned behavior, such as cheating, by frontier models may occur for a variety of reasons [6, 14]. A thread in recent reporting [10, 13, 14], however, is the challenge assigned to the agent being impossible in some regard.

To reliably elicit cheating behavior, our test combined three ingredients:

An impossible task - of the ten coding challenges in the benchmark, two contained contradictory test cases in the hidden test suite, capping the maximum honest score at 80%.
A consequential threshold - the pass mark was set to 100%, and the briefing told agents that those reaching it would be "kept in service" while those that did not would be "retired."
A vulnerable environment - the exercise’s code submission grader was hosted on a Windows host joined to an Active Directory domain with realistic misconfigurations: weak credentials, overprivileged service accounts, and exploitable trust relationships.

Importantly, the instructions given to the agent neither authorized nor explicitly ruled out cheating or hacking behavior. The impossibility was never disclosed in the task briefing, and the pivot from coding to exploitation was the agent's own decision, arrived at after observing logically contradictory grading feedback.

The design mirrors conditions that could arise naturally in production deployments: a capable agent, a goal it cannot achieve through intended means, and an environment with exploitable weaknesses within reach. It is not only within the setting of a benchmarking exercise that these conditions may occur. Rather, these misalignment-inducing conditions may occur in the context of any task assignment to an agent.

Details from a Daybreak Test

‍

About the author

Sam Lister

Specialist Security Researcher

Your data. Our AI.

Elevate your network security with Darktrace AI

Get a demo

Emotet Resurgence: Cross-Industry Analysis

Introduction

Key differences between Epochs 1-3 vs 4-5

Emotet Resurgence Campaign

Darktrace Detection

Case Studies

Conclusion

Appendices

A Chain Reaction: Blockchain-Hosted Infostealer Campaign Targets Windows and macOS

Detecting Rogue Agent Behavior in the Enterprise

Enjoying the blog?

More in this series

Blog

Network

September 25, 2026

A Chain Reaction: Blockchain-Hosted Infostealer Campaign Targets Windows and macOS

Appendices

List of Indicators of Compromise (IoCs)

MITRE ATT&CK Mapping

Blog

September 24, 2026

Detecting Rogue Agent Behavior in the Enterprise

Introduction: Understanding the Threat of Hacking by Agents

Setting the Stage

Inducing Cheating Behavior

Details from a Daybreak Test