October 7, 2026, (Inside AI) — OpenAI on Tuesday released 722 manuscripts containing 372 groups of mathematical results, all generated by an unreleased internal frontier model. The company says the output includes solutions or significant progress on hundreds of open problems in mathematics and theoretical computer science. The results are available in a public GitHub repository with protocols for revisions and citations.
The move follows weeks of intense debate after OpenAI claimed its model had resolved the Navier-Stokes existence and smoothness problem, one of the seven Millennium Prize Problems. That claim drew both excitement and sharp criticism from mathematicians who questioned how AI-generated proofs should be verified and credited.
Unlike the Navier-Stokes effort, which reportedly involved a large swarm of AI agents and millions of dollars in computing resources, nearly all of the newly released results came from a single AI agent responding to a single prompt. The model was given about 4,000 mathematical problems, with the average result requiring computing equivalent to roughly three hours of ChatGPT Pro thinking. OpenAI also published 10 summaries showing how the model approached selected problems.
Many of the proofs have been formalised using Lean, a programming language that allows mathematical proofs to be checked by computers. OpenAI said it will continue adding more formalised proofs to the repository. This formalisation step is crucial because it allows independent verification without relying on the company's own assessment.
Read: AI Model Claude Fable 5 Finds Counterexample to 1939 Math Conjecture
To address concerns about transparency and credit, OpenAI has been consulting an independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study. The group is helping develop best practices for releasing AI-generated mathematical research. OpenAI said its latest release was informed by the group's recommendations and includes protocols for revising and citing the papers.
The scale of the release has renewed questions about how AI-generated mathematics should be validated and how previous human work should be acknowledged. Researchers have called for greater transparency, independent verification, and proper attribution of prior contributions.
OpenAI said it plans to continue testing its most advanced models on mathematics and other scientific problems, arguing that AI could eventually give researchers new tools for making discoveries. The company has not disclosed which specific problems were solved or the full computational cost of the effort.
