OpenAI says an AI system solved the Navier-Stokes problem. It is one of seven Millennium Prize Problems, math’s toughest open questions.
The company said 10,000 AI agents worked together for 88 hours to reach the proof. NYU professor Tristan Buckmaster disputes the credit.
What OpenAI Claims
The Navier-Stokes equations describe how fluids move, such as air or water.
OpenAI’s proof suggests the equations can reach finite-time blowup, when fluid speeds turn infinite instead of staying smooth.
The proof came from an unreleased model that OpenAI says outperforms its released GPT-6 Astra system.
GPT-6 Astra then spent 17 hours verifying the proof’s logic.
The Clay Mathematics Institute offers $1 million for solving any of the seven problems. OpenAI said it will not claim the prize.
The Credit Dispute
NYU professor Tristan Buckmaster says he and Anthropic mathematician Levent Alpöge were close to a related result.
Their work sat inside OpenAI’s Codex, a coding-focused AI model. Buckmaster said that may have exposed it to OpenAI’s team.
Buckmaster addressed the concern directly in a public statement:
“I do not know what their model did, or how. I do not know whether our data was used. I am not accusing anyone of anything.”
— Tristan Buckmaster, NYU mathematician
OpenAI researcher Sebastien Bubeck responded at a press briefing. He denied the company had accessed or used their unpublished work.
OpenAI also acknowledged something else. It cannot rule out that the researchers’ product use shaped its model’s training.
The dispute comes as OpenAI prepares for a reported $1 trillion public listing. Anthropic is pursuing an even larger valuation through its own IPO push.