Anthropic's Claude Automates Mathematical Safety Proofs for Code

Anthropic’s Claude has demonstrated the ability to autonomously generate mathematical proofs that verify the correctness of existing software, marking a significant shift in how developers approach code safety. By partnering with Theorem, the company utilized the AI to produce approximately 180 million lines of proof code for 123 open-source libraries within a few weeks. This initiative represents a major step toward making formal verification a routine part of the software development lifecycle, similar to standard testing.

Formal verification has historically been a resource-intensive process reserved for high-stakes applications such as Airbus avionics, where the cost of an error is catastrophic. For instance, a team of specialists required 20 person-years to verify the small operating system kernel seL4. However, Anthropic’s Frontier Red Team allowed Claude to take over this labor-intensive work. The model did not write new code but instead translated crash tests from Google’s OSS-Fuzz security project into mathematical problems, analyzed compiled machine code, and generated extensive proofs. An independent tool named Rocq automatically checks these proofs, with Anthropic noting that the process requires minimal human intervention.

During the verification process, Claude identified hundreds of bugs that conventional testing methods had missed. These defects often remained hidden because they only manifested under specific conditions, such as unusually large input files or particular sequences of memory access. In one notable instance, the team found as many security bugs in a widely used PDF reader in a single week as are typically reported for that software in an entire year. Additionally, the AI uncovered a logical error in another application that had gone undetected for more than a decade, demonstrating that even failed proof attempts can yield valuable insights.

Despite the success, Anthropic emphasizes that formal verification does not equate to absolute error-free performance. The generated proofs are valid only under strict constraints, including a specific processor architecture, single-threaded execution, and input sizes under 64 MB. Since many libraries operate in varied hardware configurations and environments, these limitations mean the proofs provide a high degree of confidence rather than a universal guarantee. The company acknowledges that the underlying tools and formal foundations were built by humans at Theorem, even though the proof generation itself is described as fully autonomous.

The computational cost of this approach remains significant, with individual proofs consuming hundreds of millions of tokens. While Anthropic has not disclosed the total project costs, the potential for efficiency gains is evident. Newer models, such as Opus 5.5, are solving half the proofs with approximately 40 percent fewer tokens than their predecessors. Furthermore, a smaller test run indicated that generated proofs are reusable, which could reduce long-term expenses for developers who rely on AI-generated software.

Looking ahead, this project strengthens the argument for broader software automation. If AI can both write code and mathematically guarantee its correctness under defined conditions, the industry’s focus shifts from whether AI can program to whether it can prove its work is safe. Anthropic expects formal verification to become as practical as standard testing in the near future. Theorem’s roadmap already targets critical infrastructure components like glibc, OpenSSL, and cURL, suggesting that this technology could soon impact the foundational layers of the internet.