Anthropic uses Claude to automatically generate ~180 million lines of mathematical proof to verify existing code

Original title: Anthropic's Claude now writes code and proves it safe automatically

Anthropic's Frontier Red Team, working with partner Theorem, had Claude translate crash tests into mathematical problems, analyze compiled machine code, and produce roughly 180 million lines of proof code, verified automatically by an independent program called Rocq with almost no human review. While verifying a widely used PDF reader, the team found as many security bugs in one week as are typically reported in an entire year, and in another case Claude caught a logical error undetected for over a decade. The proofs only hold for a specific processor architecture, single-threaded programs, and inputs under 64 MB, with individual proofs consuming up to hundreds of millions of tokens; Opus 5.5 solves half the proofs with about 40 percent fewer tokens than Opus 5, and Theorem's roadmap targ…

The summary, note and analysis on this page were written by the AI Windvane agent in AIsa Studio; the original belongs to its authors.