OpenAI publicly claimed on Sept. 8 that an unreleased internal AI system solved the Navier–Stokes existence and smoothness problem, and it released a 165-page proof with a Lean formalization. [S1][S5]
Verification result: verified, because none of the weighted material claims is contradicted (0 of 450 points) and five independent reporting families directly support that OpenAI made the central claim; this verifies the announcement, not the proof's ultimate mathematical correctness. [S1][S2][S3][S4][S5]
OpenAI said up to 10,000 AI agents worked on the problem and reached the claimed solution in 88 hours. [S1][S3][S4][S5]
OpenAI said the effort began Sept. 1 after rumors that two Millennium Prize Problems had been solved. [S2][S5]
OpenAI said it completed the proof and Lean verification on Sept. 6, before publicly announcing the result on Sept. 8. [S1][S5]
Buckmaster alleged that OpenAI's parallel effort drew on work before publication, while OpenAI denied accessing that work but acknowledged that de-identified product-use data might have improved its models. [S2][S4][S5]
The supplied reporting establishes a claimed solution rather than final external validation; Timothy Gowers had not yet read the proof, so its ultimate mathematical correctness remains unresolved here. [S1][S5]
