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.