OpenAI Announces AI-Generated Solution to Navier-Stokes Problem Amid Credit Dispute
OpenAI said its AI model solved the Navier-Stokes equation, a Millennium Prize problem, but mathematicians Tristan Buckmaster and Levent Alpöge claim the company used knowledge of their parallel work and dispute credit for the breakthrough.
OpenAI announced that an internal artificial-intelligence model produced a solution to the Navier-Stokes equation, one of the seven Clay Millennium Prize problems valued at $1 million. The company said the effort began after training a new model with advanced mathematical capabilities on Aug. 28, and that more than 1,000 AI agents worked on the problem for over 50 hours before delivering a Lean-formalized proof.
The announcement has been challenged by NYU mathematician Tristan Buckmaster and Anthropic researcher Levent Alpöge, who posted documents earlier in the week describing their own progress on related equations using AI tools such as Claude and Codex. Buckmaster alleges that OpenAI learned of their work, increased resources toward the problem, and later offered to publish a paper crediting OpenAI’s model without including Alpöge’s name.
OpenAI officials denied accessing the pair’s Codex prompts. > "We, whether it’s the researchers or the agents, did not see any of their work until it was released publicly last night," said Sebastien Bubeck, a mathematician and AI researcher at OpenAI. > "I want to be extremely clear that we recognize the priority of Levent Alpöge and Tristan Buckmaster’s work on unforced Euler, and we have nothing but congratulations to them… We did not use their prompt or proof to prompt our models or direct our agents," he added.
Mark Chen, head of research at OpenAI, said the computational effort cost “in the millions of dollars.” Ven Chandrasekaran, another OpenAI mathematician, noted that the company’s solution differs substantially from the one reported by Buckmaster and Alpöge. The dispute highlights emerging tensions over attribution as AI systems become more involved in mathematical research.