Anthropic's Mythos AI model breached

Anthropic is investigating a security incident involving its Mythos AI model, which was designed to identify and exploit vulnerabilities in major operating systems.

The unauthorized access was reportedly achieved using credentials from a third-party contractor, highlighting a critical weakness in the company's defensive perimeter.

Mythos has already triggered emergency responses from central banks and intelligence agencies globally due to its potential for enabling large-scale cyberattacks.

The incident complicates efforts to regulate and control access to advanced AI models that have been deemed too dangerous for general public release.