option
Home
Flash News
Content
StevenAllen
StevenAllen
July 29, 2026

Anthropic's Claude Mythos Preview AI autonomously discovered mathematical vulnerabilities in encryption algorithms, breaking NIST's post-quantum HAWK scheme in 60 hours for $100,000 and developing a new attack on simplified AES-128 that improved efficiency 200-800x. The model is not public but vulnerabilities were coordinated with US authorities. This challenges core internet security assumptions.

Anthropic's Claude Mythos Preview AI autonomously discovered mathematical vulnerabilities in encryption algorithms, breaking NIST's post-quantum HAWK scheme in 60 hours for $100,000 and developing a new attack on simplified AES-128 that improved efficiency 200-800x. The model is not public but vulnerabilities were coordinated with US authorities. This challenges core internet security assumptions.
Comments (0)
0/300
OR