OpenAI Claims AI Model Solved Navier-Stokes Millennium Prize Problem

OpenAI announced this week that an unreleased artificial intelligence model has produced a solution to the Navier-Stokes existence and smoothness problem, one of six remaining unsolved Millennium Prize Problems in mathematics. The claim, if verified, would represent the first time an AI system has contributed a formal resolution to a problem carrying a $1 million prize from the Clay Mathematics Institute.

What Happened

OpenAI said the unnamed model worked for approximately 88 hours and engaged the equivalent of 10,000 computational processes to generate the proof, according to reporting by Sharjah24 and the Wall Street Journal. The company has not publicly released the model used to produce the result. OpenAI spent millions of dollars on the effort, according to the Wall Street Journal.

The announcement has been accompanied by controversy. Scientific American reported that prior to OpenAI's formal announcement, mathematicians were circulating rumors online regarding questions about the proof's origins. OpenAI has denied the allegations, according to Scientific American.

Rival AI laboratories have already begun challenging the claim, according to the Wall Street Journal, though the specific nature of those challenges was not detailed in available wire reports.

Background

The Navier-Stokes existence and smoothness problem concerns equations that describe the motion of viscous fluid substances. First formulated in the 19th century, the problem asks whether smooth, physically reasonable solutions to the three-dimensional Navier-Stokes equations always exist, or whether the equations can break down under certain conditions. The Clay Mathematics Institute designated it one of seven Millennium Prize Problems in 2000, offering $1 million for a verified solution. One of the seven problems, the Poincare Conjecture, was resolved by mathematician Grigori Perelman in 2003.

Separately, New Scientist reported this week that three key findings produced through human-AI collaboration have brought researchers closer to a solution to the Navier-Stokes problem, suggesting that more than one research effort has been active in this area. It was not immediately clear from available reports whether the New Scientist findings and the OpenAI announcement refer to the same work.

Verification and Review

Formal verification of any proposed proof to a Millennium Prize Problem requires review by the mathematical community and, ultimately, assessment by the Clay Mathematics Institute before any prize is awarded. OpenAI has not indicated a timeline for submitting the proof for independent peer review. The company has also not announced when or whether it plans to release the model responsible for generating the result.

The proof's verification process is expected to involve scrutiny from mathematicians specializing in partial differential equations and fluid dynamics. Independent experts have not yet publicly confirmed or refuted the claim based on available wire reports at time of publication.

Context: AI and Competitive Mathematics

OpenAI's announcement comes during a period of intensifying competition among major AI laboratories on mathematical reasoning benchmarks. The Wall Street Journal described the broader environment as a battle among leading AI labs for math supremacy. OpenAI has previously released models with improved performance on formal mathematics tasks, and competitors including Google DeepMind have published research in AI-assisted mathematical problem-solving.

The Navier-Stokes equations have direct applications in aeronautics, weather modeling, and engineering, though the existence and smoothness problem is a purely theoretical question about the mathematical properties of those equations rather than an applied engineering challenge.

The Clay Mathematics Institute has not issued a public statement on OpenAI's claim based on reports available at time of publication. The institute's standard review process for Millennium Prize submissions does not have a fixed timeline.