OpenAI Says Its AI Proved a Navier-Stokes Singularity

OpenAI announced an AI-generated proof that 3D fluid motion can blow up in finite time, but academics are scrutinizing the claim and its origins.

OpenAI announced on September 8, 2026 that an internal multi-agent AI system has proved that smooth three-dimensional fluid motion can develop a mathematical singularity in finite time with finite energy — the famous 'blow-up' question at the heart of the Navier-Stokes Existence and Smoothness problem, one of the seven Millennium Prize Problems selected by the Clay Mathematics Institute in 2000. The company released a paper titled 'Finite Time Blowup for Navier-Stokes' together with a formalization checked in the Lean proof assistant, which enforces a strict standard of formal correctness. OpenAI has already stated that it does not intend to claim the million-dollar prize attached to the problem. To date, only one Millennium Prize Problem, the Poincaré Conjecture, has been officially solved, by Grigori Perelman, who declined both the award and the prize money.

The scale of the effort was extraordinary. The result, if verified, would show that the equations used daily for more than a century in meteorology, engineering, hydraulics and fluid dynamics contain an internal contradiction: solutions starting from smooth conditions can reach infinite velocity within finite time, revealing an intrinsic limitation of the continuous mathematical model.

The announcement quickly drew a serious accusation. i-programmer.info reports that after OpenAI's announcement, New York University mathematics professor Tristan Buckmaster publicly claimed OpenAI stole unpublished research he had been developing with Levent Alpöge, a mathematician at Anthropic. According to the report, Buckmaster alleged that information about their private AI-assisted work, which they had been running through tools like Codex, was prematurely absorbed or leaked into OpenAI's training or search pipelines, enabling OpenAI to close in on the proof and announce it first. OpenAI's paper also acknowledged prior analytical foundations — the framework for singularity formation and scale amplification established by Spanish mathematicians Diego Córdoba and Luis Martínez-Zoroa — while, per i-programmer.info, failing to mention contemporaneous AI-assisted research involving Anthropic.

Verification is now the central question. While the Lean proof assistant ensures the mathematics meets a strict standard of formal correctness, i-programmer.info notes the academic community may not be easily satisfied about the work's conceptual validity, and the claim is under intense academic scrutiny. yourweather.co.uk adds a scope caveat: skeptics point out the solution applies only to the case involving 'external forcing' — a force from outside the system acting on the fluid — while the problem remains unsolved for an isolated fluid. The outlet also notes that if confirmed, the discovery would be primarily conceptual, establishing a theoretical limit for classical fluid dynamics: at molecular scales, molecular collisions, thermodynamic effects and quantum fluctuations dissipate energy in ways the Navier-Stokes equations do not account for. Practical benefits, such as refinements to simulation software and high-resolution weather models, would likely be indirect and long-term, and further confirmation and explanation are still needed.