Anthropic Claude verifies open-source code with 180 million lines of proof

Anthropic used its Claude model to mathematically verify the correctness of existing software code.

The Frontier Red Team at Anthropic partnered with Theorem to perform formal verification on 123 open-source libraries from Google's OSS-Fuzz project. In a matter of weeks, Claude generated approximately 180 million lines of mathematical proof code. This output was used to analyze compiled machine code and turn crash tests into mathematical problems. A separate verification program called Rocq automatically checks these proofs, requiring almost no human reviewers.

Traditional formal verification is expensive and slow. Specialists required 20 person-years to verify seL4, a small operating system kernel. This high cost has limited the technique to critical systems like Airbus avionics software. Claude did not write new software but verified existing code. The researchers describe the proof generation as fully autonomous, although humans at Theorem built the underlying tools and formal foundations.

During the verification process, Claude encountered hundreds of bugs and was permitted to fix them. Many of these errors would likely have escaped conventional testing because they only appear under narrow conditions, such as unusually large input files or specific access sequences. When verifying a widely used PDF reader, the team found as many security bugs in one week as are usually reported for that software in a full year. In another application, Claude identified a logical error that had gone unnoticed for more than a decade. Anthropic notes that even failed proof attempts can reveal real bugs.

The company states that formally verified does not mean error-free. The proofs hold only under defined conditions: a particular processor architecture, single-threaded programs, and inputs smaller than 64 MB. Real-world deployments often run on different hardware and configurations. Individual proofs consume hundreds of millions of tokens, and Anthropic has not published total project costs.

Opus 5.5 solves half the proofs using about 40 percent fewer tokens than Opus 5, indicating improving efficiency. Theorem's roadmap targets critical infrastructure including glibc, OpenSSL, and cURL. Anthropic argues that if AI can verify its own code, developers can trust AI-generated software more and spend less time reviewing it, cutting costs across the coding process.

The project shifts the focus from whether AI can program to whether it can prove the code it writes is safe. Anthropic expects formal verification with AI to become as practical as standard testing.