OpenAI uploads 722 AI math preprints; three pulled next day

OpenAI uploaded 722 AI-generated mathematics preprints to GitHub on 6 October 2026, claiming progress on 372 open problems in geometry, algebra and computer science. Nature reports that the manuscripts were produced by an unreleased model, and that a day after the upload the company withdrew three, after a sign error invalidated the argument in one paper and the constructions in two others that depended on it. None of the 722 papers claimed a solution to any of the Clay Mathematics Institute's five remaining Millennium Prize Problems.

Alvaro Lozano-Robledo, a number theorist at the University of Connecticut, called the volume staggering. "Why is OpenAI trying to solve hundreds of math problems to be released all at once?" he asked.

Levent Alpöge, a mathematician at rival lab Anthropic, acknowledged there would be "sad stories" for researchers whose active projects had been pre-empted, but called the release "obviously the most significant moment in mathematical history."

Others were less generous. Ursula Martin at the University of Oxford said OpenAI was "tossing the community a messy first draft and expecting us to clean it up for them." Michael Harris, a number theorist at Columbia University, refused to engage at all. "I can't bring myself to spend my time studying the outcome of this process, which I consider illegitimate," he said. "So I will not be reading these papers."

An OpenAI spokesperson said the company's goal was to provide tools for mathematicians and scientists, and to evaluate its models on difficult frontier problems.