OpenAI has announced that it used an internal, unreleased AI model to solve the Navier-Stokes problem, one of seven Millennium Prize Problems that has stood unsolved for roughly 90 years and relates to the flow of liquids and gases. The claimed breakthrough, revealed in a blog post, is significant for the mathematics community, but has been overshadowed by a dispute over whether OpenAI improperly drew on research data from rival mathematicians without proper acknowledgement.

The controversy centres on NYU professor Tristan Buckmaster, who published related findings with Anthropic researcher Levent Alpöge just a day before OpenAI's announcement, using a similar approach developed with OpenAI's Codex and Anthropic's Claude. Buckmaster says he pressed OpenAI on whether its model had trained on or accessed their private Codex sessions, but received evasive answers. OpenAI maintains no specific user data was accessed, though it concedes it cannot fully rule out that de-identified usage data indirectly aided its models; OpenAI staff member Sebastien Bubeck said the two proofs differ significantly and that OpenAI had not seen Buckmaster's work before it was published. OpenAI says it does not intend to claim the associated $1 million prize.