Anthropic investigates report of rogue access to hack-enabling Mythos AI

The Guardian — Apr 22, 08:58 AM

‘Handful’ of people allegedly gain unauthorised access to model adept at detecting cybersecurity vulnerabilities

The AI developer Anthropic has confirmed it is investigating a report that unauthorised users have gained access to its Mythos model, which it has warned https://www.theguardian.com/technology/2026/apr/08/anthropic-ai-cybersecurity-software">poses risks to cybersecurity.

The US startup made the statement after Bloomberg reported on Wednesday that a small group of people had accessed the model, which has not been released to the public because of its https://www.theguardian.com/business/2026/apr/13/goldman-sachs-chief-hyper-aware-risks-anthropics-mythos-ai-david-solomon">ability to enable cyber-attacks.