OpenAI claims it solved Navier-Stokes, and mathematicians aren't buying it

The company said an internal model stronger than GPT-6 Astra, running with 10,000 agents in parallel, had cracked one of the seven Millennium Prize problems that has remained open for roughly ninety years. The cash award is one million dollars, but OpenAI has already made clear it does not intend to collect it.

The timing raises questions

A day before OpenAI's announcement, New York University professor Tristan Buckmaster and Anthropic researcher Levent Alpöge published findings on a closely related problem. Buckmaster says he reached out to the company after learning it was aware of their progress, and discovered that OpenAI's proof follows a similar path to the one they developed — using OpenAI's Codex and Anthropic's Claude.

User data or coincidence

Buckmaster asked directly whether the model had been trained on their Codex sessions, where every project draft was saved. The answer he received: the model did not "peek" at user data. When he pressed on the training question, he got no reply. In a public statement OpenAI said it "did not access specific user data to solve the problem," but added that it "cannot rule out that semi-identified data originating from their use of our products helped improve the models."

Contrad