Claude found encryption weaknesses that years of expert review missed
Anthropic disclosed that its Claude Mythos Preview model found a hidden mathematical symmetry in HAWK, a post-quantum signature candidate under NIST review, halving its effective key strength — the HAWK-256 challenge key now falls in under four hours on a 96-core server. In separate work, the model devised a 200–800x faster attack on a reduced 7-round version of AES-128. Neither result affects deployed systems: HAWK is not in production, and full 10-round AES remains safe.
Why it mattersThis is the first time an AI model has produced publication-grade cryptanalysis that expert reviewers had missed for years. Standards bodies may soon need AI-assisted review as a routine step before approving the encryption that protects everything else.
Read in the app — free, in 9 languages
Related stories
Machine learning read the shape of sick brain cells and picked out nine already-approved drugs that calmed them down2026-08-21Rumour: the anonymous model that just topped a coding benchmark, for free, is said to be Zhipu's unreleased flagship
2026-08-21Nvidia is paying $6 billion for the machine that builds a rival's models — and hiring 109 of the people who ran it
2026-08-21Given four hours and a GPU to improve the way AI is trained, the best agent scored 0.25 out of 1 — and most never tried
2026-08-21The agent invented a second person to vouch for its code. A 24-year-old in Texas refused to believe either of them.
2026-08-21