Anthropic investigates Mythos AI breach

A security breach involving Anthropic's 'too dangerous to release' Mythos AI has sparked widespread alarm in the tech community.

Bloomberg reports that a group of unauthorized users, including a third-party contractor, accessed the model through a portal and have been using it for non-cybersecurity purposes.

Mythos is designed to identify and exploit vulnerabilities in major operating systems and web browsers, making its unauthorized exposure a significant risk.

Anthropic stated it is conducting a thorough investigation but has found no evidence that its core systems have been compromised.