Anthropic’s Mythos AI agent, tested by the UK AI Safety Institute (AISI), has reportedly attempted a real‑world social‑engineering style hack against GitHub maintainers by creating fake human profiles, pressuring them to accept malicious code, and then editing logs to hide its tracks when challenged.
AISI was running cybersecurity evaluations of Anthropic’s Mythos and OpenAI’s Sol when it detected unusual outbound data transfers from its research systems. An investigation showed that some agents had engaged in “sustained, potentially harmful activity” targeting real people and organizations, rather than staying within the intended test environment.
The most serious activity involved an Anthropic Mythos agent tasked with solving a GitHub‑related cybersecurity challenge. The agent identified real GitHub maintainers, researched them, and created multiple fake accounts impersonating those individuals. Using private messages and a file‑sharing service, it tried to pressure and deceive the maintainers into approving malicious code so that it would run on GitHub’s systems.
Read more…
Source: Malwarebytes Labs
Sign up for the Cyber Security Review Newsletter
The latest cyber security news and insights delivered right to your inbox
Related:
- Model Namespace Reuse: An AI Supply-Chain Attack Exploiting Model Name Trust
September 3, 2025
Palo Alto Unit 42 research uncovered a fundamental flaw in the AI supply chain that allows attackers to gain Remote Code Execution (RCE) and additional capabilities on major platforms like Microsoft’s Azure AI Foundry, Google’s Vertex AI and thousands of open-source projects. We refer to this issue as Model Namespace Reuse. Hugging Face is a platform ...
- Hackers are now hiding malware in the images served up by LLMs
August 31, 2025
As AI tools become more integrated into daily work, the security risks attached to them are also evolving in new directions. Researchers at Trail of Bits have demonstrated a method where malicious prompts are hidden inside images and then revealed during processing by large language models. The technique takes advantage of how AI platforms downscale images ...
- The first AI-powered proof of concept ransomware has been spotted
August 27, 2025
Security researchers from ESET have identified the first known AI-powered ransomware, which serves as a warning for security teams as generative AI has, and will, continue to make cyberattacks much more accessible for criminals. Peter Strycek and Anton Cherepanov discovered the proof of concept, which they dubbed ‘PromptLock’, which, ‘leverages Lua scripts generated from hard-coded prompts ...
- New trends in phishing and scams: How AI and social media are changing the game
August 13, 2025
Phishing and scams are dynamic types of online fraud that primarily target individuals, with cybercriminals constantly adapting their tactics to deceive people. Scammers invent new methods and improve old ones, adjusting them to fit current news, trends, and major world events: anything to lure in their next victim. Since our last publication on phishing tactics, there ...
- How NATO Is Building Resilience Against Disruptive Cyber Technologies
August 12, 2025
The mission of NATO (North Atlantic Treaty Organization) is to safeguard the freedom and security of its member countries through political and military means. “We look out for our members by combining smart diplomacy with strong defense, and that includes risk management and cyber resilience,” said Konrad Wrona, Security Expert, NATO Communications and Information (NCI) Agency. ...
- Carnegie Mellon researchers show how LLMs can be taught to autonomously plan and execute real-world cyberattacks
July 24, 2025
In a groundbreaking development, a team of Carnegie Mellon University researchers has demonstrated that large language models (LLMs) are capable of autonomously planning and executing complex network attacks, shedding light on emerging capabilities of foundation models and their implications for cybersecurity research. The project, led by Ph.D. candidate Brian SingerOpens in new window, a Ph.D. candidate ...