Anthropic's Mythos AI model sparks fears of turbocharged hacking

Ars Technica AI
Generative AI

Cyberdefenses could be exposed faster than fixes could be deployed.