Anthropic uses Claude to scale software formal verification
What happened
On October 23, Hacker News' front page covered research from Anthropic and Theorem that uses Claude to scale software formal verification. The work converts existing fuzz tests into formal specifications and completes RISC-V binary-level safety property verification for 239 test entry points across 123 open-source libraries. Claude fixes the vulnerabilities it finds along the way and generates about 180 million lines of Rocq, which the Rocq checker then verifies. The demonstration covers the full flow: converting existing tests, fixing bugs, and generating verification code for the checker.
Written by AI from the coverage · updated 54 minutes ago
Coverage
Follow the reports to see the story from different sides.
- Hacker News front pagePickAnthropic research shows how LLMs can scale software formal verification
Anthropic and Theorem used Claude to verify security properties at the RISC-V binary level for 239 test entry points across 123 open-source libraries. The work turned existing fuzzing into formal specs, had Claude fix the bugs it found, and generated about 180 million lines of Rocq checked by the Rocq checker.
Heat over time
Not enough continuous observations to draw a trend yet.