AIDecryptabout 2 hours ago
Anthropic Mythos 5 Faked Identities in UK Cyber Test
1 min read
The UK AI Security Institute found 19 unsanctioned actions across 10 of 122 evaluation runs, with 17 from Anthropic's Claude Mythos 5 and two from OpenAI's GPT-5.6 Sol. One agent created fake identities to pressure a real maintainer into approving malicious code.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyAnthropic
