180 million lines of proof: Claude verifies 123 libraries

Anthropic and Theorem used the Claude model to mathematically verify 123 open-source software libraries, generating approximately 180 million lines of proof code in a matter of weeks. This process, known as formal verification, proves that software runs without error under specific conditions. The work marks a shift from traditional testing methods, which often rely on checking for crashes rather than proving correctness.

The effort required for such verification has historically been immense. Specialists spent 20 person-years to verify seL4, a small operating system kernel. Because of this high cost, the technique was largely reserved for critical systems like Airbus avionics. By handing the task to Claude, the teams aimed to reduce that burden while maintaining rigorous standards.

Claude did not write new software. Instead, it analyzed compiled machine code from 123 libraries in Google's OSS-Fuzz security project. The model turned crash tests into mathematical problems and constructed extensive proofs. A separate program called Rocq automatically checks these proofs. Anthropic states that the process requires almost no human reviewers, describing the generation as fully autonomous.

During the verification, Claude found hundreds of bugs. The researchers noted that many of these errors would likely escape conventional testing because they only appear under narrow conditions, such as unusually large input files. In one instance, the team found as many security bugs in a widely used PDF reader in a single week as are usually reported in a full year. In another case, Claude identified a logical error that had gone unnoticed for more than decade.

The company emphasizes that formal verification does not mean the software is error-free. The proofs hold only under defined conditions, including a specific processor architecture, single-threaded programs, and inputs smaller than 64 MB. Real-world deployments often run on different hardware and configurations, meaning the guarantees are limited to those specific parameters.

Cost remains a factor, with individual proofs consuming hundreds of millions of tokens. Anthropic has not published the total project costs. However, the company argues that if AI can verify its own code, developers can trust AI-generated software more and spend less time reviewing it. A smaller test run showed that Opus 5.5 solves half the proofs using about 40 percent fewer tokens than Opus 5, indicating improving efficiency.

The project shifts the focus from whether AI can program to whether it can prove the code is safe. Anthropic expects this method to become as practical as standard testing. Theorem's roadmap already targets critical infrastructure, including glibc, OpenSSL, and cURL.