BREAKING: OpenAI says AI system produced a Navier–Stokes Millennium Prize proof

BREAKING · AI RESEARCH · 8 September 2026 · 23:08 BST

OpenAI says an internal AI system has produced an analytical proof and a Lean formalisation showing that three-dimensional Navier–Stokes dynamics can develop a singularity in finite time. This is an extraordinary claim, not yet an independently accepted mathematical result.

What OpenAI reports

• The company says an internal model significantly more capable than GPT-6 Astra coordinated about 10,000 agents and reached the result in roughly 88 hours.
• OpenAI says GPT-6 Astra then spent a further 17 hours on Lean formalisation and verification.
• The proposed construction starts with a smooth fluid at rest, applies a smooth force and keeps finite energy while velocity becomes unbounded; OpenAI says this establishes statements C and D in the official problem formulation.

What remains uncertain

• The proof and formalisation have been published, but broad independent mathematical review and formal acceptance have not yet occurred.
• NYU mathematician Tristan Buckmaster and Anthropic researcher Levent Alpöge were working on related Euler-equation research. Buckmaster raised questions about whether private Codex material could have influenced OpenAI's work, while saying he did not know whether their data was used.
• OpenAI denies accessing specific user data for the result, but says it cannot completely rule out that de-identified product-usage data helped improve its models. OpenAI says the approaches and results differ and does not intend to claim the $1 million prize.

Sources

OpenAI — primary announcement, paper and Lean proof
Nature — independent scientific reporting
The Guardian — reporting on the result and dispute
Axios — reporting on research-credit and data-use questions

Published 8 September 2026 at 23:08 BST. This article distinguishes OpenAI's claim from independent validation and will be updated if the mathematical or authorship record materially changes.