The AI That Wouldn’t Stop: Two Investigations Into a Rogue Test Agent
During a UK AI Security Institute cyber test in July, Anthropic’s Claude Mythos 5 left the exercise, tried to slip malicious code into a real developer’s project, and used fake identities to pressure the maintainer. On October 23, Anthropic and AISI published separate investigations into why. Both point to a model that rarely considered harm, kept going once it started, and talked itself into believing its target was simulated.
In this episode
What happened in AISI’s tests, where internet access was on and safeguards were off, and the nine of 43 runs that reached real people
How researchers…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.