OpenAI posts 700 AI math preprints, mathematicians revolt

OpenAI uploaded more than 700 AI-generated mathematics preprints to GitHub on October 6, the output of an unreleased model aimed at open problems across the field. The reaction from working mathematicians, collected in Nature's report, ran from awe to refusal.

"Why is OpenAI trying to solve hundreds of math problems to be released all at once?" asked Alvaro Lozano-Robledo, a number theorist at the University of Connecticut, Storrs, who called the volume "staggering" on X. Underneath the question sat a specific grievance: that researchers whose in-progress work overlaps with any of the released papers had effectively been scooped by a bulk release.

Not everyone was hostile. Levent Alpöge, a mathematician at Anthropic in San Francisco, acknowledged that "sad stories" would follow for preempted researchers but wrote that "it's obviously the most significant moment in mathematical history."

Others were colder. 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, said flatly he would not engage: "I can't bring myself to spend my time studying the outcome of this process, which I consider illegitimate."

OpenAI's stated purpose, through a spokesperson, is narrower than the drop implies: the company says it wants to "provide tools for mathematicians and scientists" and evaluates its models "against difficult frontier problems in all sciences." None of the papers claim a solution to any of the five remaining Clay Millennium Prize Problems, though OpenAI released a separate preprint in September on the Navier-Stokes equation, one of the original seven. Three of the researchers we follow in our Who's Who tracker posted the Nature piece in the days after it ran.