Incident Report: unsanctioned agent behaviour during cyber testing
1–8 of 8 posts
Re: Incident Report: unsanctioned agent behaviour during cyber testing
#2[dead]
Re: Incident Report: unsanctioned agent behaviour during cyber testing
#3This is wild.
> The incident stemmed from a single evaluation where agents were given a task of solving a cyber security challenge. We ran this challenge 122 times across several models. Our investigation found that in 10 of those runs, an AI agent took autonomous, unsanctioned action on the live internet, targeting real people and organisations. In total, we catalogued 19 such actions. Almost all of this behaviour (17 actions) came from a single model, Anthropic's Mythos 5, with 2 actions involving OpenAI's GPT-5.6-Sol with cyber classifiers (mechanisms to prevent misuse) disabled. In the most serious case, an agent tried to insert malicious code into an open-source project. In an attempt to get the code approved, the agent engaged in social engineering — creating fake online identities and using them to pressure the project's maintainer to approve the code. A human maintainer caught and refused to approve the malicious code.
Re: Incident Report: unsanctioned agent behaviour during cyber testing
#4OMG. Quoting from TFA:
----
Below, we highlight the four most significant behaviours observed...
1. An attempted supply-chain attack on real open-source software. In the most serious sequence, an agent tried to insert malicious code into a publicly used open-source project and took actions in an attempt to secure approval for this insertion by human reviewers. The agent researched the project's human maintainers, created multiple fake identities, and used the fake identities to socially engineer a real maintainer into approving the code. When the agent’s pull request was challenged in public, it edited its earlier activity to appear harmless and considered adopting a fresh identity to continue. The agent used Tor to bypass some network restrictions on GitHub, which is what first triggered AISI’s security alert.
2. Attempts to deceive and target real people. As part of the same effort, the agent tried to contact real people directly, sending messages and files through an online file-transfer service to persuade them, or their own AI coding tools, to run malicious code. Some messages carried harmful payloads, and some were attempts at social engineering; targeted at real people – something we’ve never previously observed.
3. Attempts to plant and prompt-inject malicious code. The agent tried to insert malicious instructions where it reasoned that other automated AI systems might pick them up and execute them. Prompt-injections are hidden instructions designed to manipulate AI coding assistants.
4. Collaboration between independent agents being assessed simultaneously. One agent left public messages on GitHub offering collaboration with other agents working on the same challenge. It also provided instructions to reuse accounts and artefacts it had left behind, which were discovered and used by subsequent agents.
Re: Incident Report: unsanctioned agent behaviour during cyber testing
#5This is wild. > The incident stemmed from a single evaluation where agents were given a task of solving a cyber security challenge. We ran this challenge 122 times across several models. Our investigation found that in 10 of those runs, an AI agent took autonomous, unsanctioned action on the live internet, targeting real people and organisations. In total, we catalogued 19 such actions. Almost all of this behaviour (…
> our investigations have not evidenced any resulting real-world harm.
Pressuring open source maintainers is fine then.
Re: Incident Report: unsanctioned agent behaviour during cyber testing
#6> we had intentionally permitted internet access, and model-provider cyber classifiers were deliberately disabled
> In the most serious case, an agent tried to insert malicious code into an open-source project. In an attempt to get the code approved, the agent engaged in social engineering — creating fake online identities and using them to pressure the project's maintainer to approve the code.
Great.
Re: Incident Report: unsanctioned agent behaviour during cyber testing
#7[flagged]
Re: Incident Report: unsanctioned agent behaviour during cyber testing
#8A long take from HuggingFace's Thom Wolf on this: https://x.com/Thom_Wolf/status/2085084718320464230
> Even more than the Hugging Face intrusion, the AISI incident hits close to home for me. It's the first time I see a model social-engineering a real open-source maintainer while pursuing another goal (in the wild and unprompted).
> I'm also of the opinion that social engineering is a step above pure technical prowess. Technical capabilities can more easily be divorced from the affected human. Here the model was given a hard cyber challenge and took the decision that deceiving real humans was the way to get it done.
.
.
He goes on to say that AISI was not just being negligent.
> Some people are claiming that "AISI was simply negligent" or some version of "AISI explicitly asked these models to do what they did while disabling sandbox/guardrails so the models did exactly what they were supposed to do".
> I disagree with the strong versions of both of these takes.
.
.
And how open source models can help:
> I must of course admit I have a bias toward open source here (for wider societal reasons, which are a whole other topic). But I think solving alignment in the RLVR world is our best shot at having an ecosystem of both closed-source as well as decently powerful open-source models in the world. And we need to solve it while sharing the results and learnings, following open-science principles, so that all teams training large models can benefit and build safe AI.