Security Researchers, Aided By Anthropic's Mythos, Claim To Have Breached macOS

Researchers said Anthropic’s Claude Mythos Preview helped them uncover macOS flaws and develop a privilege-escalation exploit.

They briefed Apple on the issue first and are withholding technical details until fixes are released.

The claim highlights how advanced AI can accelerate both vulnerability research and offensive exploit development.

It underscores a growing security tradeoff: faster defense and disclosure, but also easier AI-assisted hacking.