Anthropic's Mythos AI raises alarm

Anthropic's powerful new AI model, Mythos, has sparked urgent security talks among financial regulators and national governments due to its potential hacking capabilities.

The company has withheld a full public release, warning that the model could be used to identify and exploit software vulnerabilities.

US and UK officials are bracing for a new era of cyber risk as AI models demonstrate the ability to find bugs with frightening speed.

While some firms are testing the tool to improve defenses, others fear it could usher in a 'Bugmageddon' for critical infrastructure.