OpenAI Releases 372 AI-Generated Math Proofs on GitHub

A Massive Dump of AI Math

OpenAI has released 372 mathematical results generated by an internal frontier model, posting them all on GitHub instead of submitting them to peer-reviewed journals. The collection includes improvements to major computer algorithms and advances related to the Riemann hypothesis, a famous unsolved problem in mathematics ¹.

Most of the results came from a single prompt given to a single AI agent, with each one taking about three hours of ChatGPT Pro compute on average. This is a sharp contrast to the company's earlier Navier-Stokes solution, which required a swarm of 10,000 agents and millions of dollars in compute ¹.

Why GitHub Instead of Journals?

OpenAI made a deliberate statement by skipping traditional academic journals. The sheer volume of AI-generated results could overwhelm the math community's capacity for manual review, and the company argued that the traditional process of doing science is simply too slow for this pace of discovery ¹.

Many of the proofs come with formalizations in Lean, a programming language built for machine-checkable mathematical proofs. More formalizations are planned. Formal verification can confirm that a proof is logically correct, though it cannot judge whether a result is mathematically relevant or original ¹.

Outside Advice, But on Their Terms

Before the release, OpenAI consulted the Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study in Princeton. The group includes Fields Medal winner Timothy Gowers and other renowned mathematicians. However, the company set one clear boundary: the mathematicians could advise on how results are communicated, but not on whether or how fast they are produced ^{1 2}.

OpenAI also announced it will fund workshops and conferences focused on helping mathematicians understand and vet the AI-produced results ¹.

The Math Community Is Divided

Reactions range from excitement to frustration. In an open letter titled "A Severe Misalignment of AI in Mathematics," 25 Fields Medal winners warned of a deep disconnect between the AI industry's goals and those of mathematics. They argued that problem-solving is merely a tool, and that the real goal of math is conceptual understanding and insight ¹.

Timothy Gowers has warned that within one to two decades, mathematical literature could grow enormously while no human community remains that truly understands it. Fields Medal winner Terence Tao said that training young mathematicians needs to emphasize the human side and tightly limit AI tool use so genuine learning survives ¹.

OpenAI acknowledged it wants to improve the quality of its citations and presentation, and says it is working on a responsible release of the model to "directly empower scientists with state-of-the-art capabilities" ¹.

Key facts

• OpenAI released 372 AI-generated math results on GitHub, bypassing traditional journals
• Most results came from a single prompt to a single AI agent, averaging three hours of ChatGPT Pro compute each
• Proofs include formalizations in Lean for machine-checkable verification
• 25 Fields Medal winners signed an open letter warning about AI's misalignment with mathematical goals
• OpenAI will fund workshops and consult the Institute for Advanced Study's advisory group

Sources

1. OpenAI dumps 372 AI-generated math proofs on GitHub, telling the academic world to keep up — the-decoder.com
2. OpenAI Funds Math Workshops to Vet 372 AI Proofs 2026 - shattered.io — shattered.io