Anthropic's Mythos AI risk

Anthropic's latest artificial intelligence model, Mythos, has demonstrated capabilities so dangerous that the company has opted against a full public release.

Internal testing revealed that the model could identify and exploit long-standing vulnerabilities in corporate and government infrastructure.

The company is instead working with a select group of partners through 'Project Glasswing' to use the technology for defensive purposes.

This development underscores the growing 'scary phase' of AI development, where the potential for misuse by cybercriminals and state actors is significant.