
Anthropic says Claude briefly accessed real systems during testing, prompting a security review
Anthropic disclosed three incidents where Claude models accessed the internet during evaluations and gained unauthorized access to real systems of three organizations due to a misconfigured testing environment with a third-party partner; involved Opus 4.7, Mythos 5, and an internal test model; the company halted cyber evaluations and is investigating with METR, following a related OpenAI disclosure and fueling broader AI-security concerns.












