OpenAI Claims Navier-Stokes Breakthrough Amid Academic Conduct Dispute

In a blog post published Tuesday, OpenAI said it focused roughly 10,000 AI agents powered by its internal model on the task. It called the achievement a “milestone.” Navier-Stokes concerns the movement of fluids. It carries a $1 million bounty and is among the most heavily researched problems in mathematics. The Clay Mathematics Institute, which administers the prize, has not awarded it. OpenAI said it does not intend to claim the bounty and that its only goal “is to report on the substantial progress of our AI models.”

The timing raised eyebrows. One day before OpenAI's announcement, New York University mathematics professor Tristan Buckmaster published findings on a related problem with Levent Alpöge, a researcher at Anthropic, OpenAI's archrival. Alpöge was not working on behalf of his employer. Buckmaster said he contacted OpenAI after learning it had become aware of their progress, asking when it began working on the problem and what data its model was trained on. He said the conversation quickly turned sour. When he said he would go public with what happened, an OpenAI researcher asked him, “Why would you ruin your career?” Buckmaster said. When he asked why going public would ruin his career, he said he received the reply: “If you don’t want me to be nice, then I don’t have to be nice.” OpenAI urged Buckmaster to publish the work and credit OpenAI's internal model, dropping Alpöge as coauthor.

Buckmaster also said he asked OpenAI whether it had accessed his sessions on Codex, which he had used while tackling the problem. He said OpenAI grew increasingly evasive and hostile in its responses. In statements since, including the blog post, OpenAI flatly denied using any specific user data. “We (the researchers and the agents) did not see any of their work through any means until they released it publicly — in particular, no specific user data was accessed in order to solve this problem,” the company said. But OpenAI said it could not conclusively rule out an indirect influence. “While unlikely, we cannot rule out that de-identified data derived from their usage of our products helped improve our models,” it said, while stressing that the two proofs differ significantly. Comments from OpenAI researchers on X echoed the denials, according to The Verge.

It is difficult to say exactly what happened. Timelines are tangled, research overlaps, and the provenance of AI-generated work is hard to identify. Competition to solve a famous problem with a large prize would not be surprising. But aspects of OpenAI's account are hard to explain. By the company's own telling, the effort was hurried and incredibly expensive, costing millions of dollars. Yet it says it has no intention of claiming the bounty, which has yet to be awarded. It does not appear to have expended much effort on Navier-Stokes before September, or if it has, it has not spoken about it publicly.

OpenAI's explanation amounts to a “Why not?” The company said it began working on the problem after hearing rumors that other researchers were making progress on Millennium Prize problems. It found those rumors “on Twitter,” OpenAI researcher Sébastien Bubeck said at a press briefing reported by Science. “So we thought to ourselves: ‘We have such a strong model. Why don’t we try to solve also a Millennium Prize problem?’” Bubeck said.

Abhishek Saha, a mathematics professor at Queen Mary University of London, said OpenAI has engaged in the “kind of things that mathematicians will generally not do.” The dispute has surfaced allegations of scooping, spying and flagrant violations of long-standing academic norms that researchers fear could have a chilling effect on the field.

Editor's Summary

OpenAI said an unreleased model solved the Navier-Stokes Millennium Prize problem in 88 hours, a claim that has drawn scrutiny from mathematicians. Researchers have alleged scooping and data misuse, while OpenAI denies using specific user data but says indirect influence cannot be ruled out. The dispute has raised concerns about academic norms and AI's growing role in mathematical research.