Anthropic investigating claim of unauthorised access to Mythos AI tool
Anthropic is investigating a claim that a small group of people gained unauthorized access to its Claude Mythos model, a powerful AI tool used for cyber-security, according to a report from the BBC.