San Francisco, California – September 8, 2026
OpenAI has announced that an internal AI system has resolved one of mathematics' most enduring open problems: the Navier-Stokes existence and smoothness problem. The company says its swarm of approximately 10,000 AI agents working in parallel produced a proof showing that three-dimensional fluid flow equations can develop a singularity in finite time.
The Navier-Stokes equations, formulated in the 1800s, describe how fluids such as water and air move through space. They underpin calculations ranging from aircraft design to weather forecasting. For nearly a century, mathematicians have asked whether solutions to these equations always remain well-behaved or can sometimes "blow up"—developing infinite velocity in finite time.
An 88-Hour Computational Sprint
OpenAI's effort began on September 1, when researchers heard rumors that independent teams had made progress on related fluid dynamics problems. The company launched a coordinated attack using its internal next-generation model, which it describes as significantly more capable than GPT-6 Astra.
By September 5—about 88 hours after the first agents were deployed—the system had constructed a solution demonstrating finite-time singularity formation. The proof consists of a 166-page analytical writeup and a formalization in Lean, a programming language used to verify mathematical proofs.
The Vortex That Breaks the Equations
The solution OpenAI found is a vortex—a spinning swirl of fluid—that spirals inward and becomes increasingly elongated. In this mathematical scenario, the fluid starts at rest under a smooth external force. As time progresses, the vortex concentrates its motion into a narrowing region while maintaining finite kinetic energy, yet the velocity at its core grows without bound.
This counterexample establishes what mathematicians call "statement C" of the Clay Mathematics Institute's official Millennium Prize formulation. It shows that smooth solutions to the Navier-Stokes equations do not exist for all time under certain physically reasonable conditions.
"Our proof does show that there exist fluids which start out perfectly normal, and under the Navier-Stokes equations, actually achieve infinite speed in a finite amount of time," said OpenAI computer scientist Ven Chandrasekaran in a press briefing. Because real physical fluids cannot move at infinite speed, the result suggests the equations may have limited applicability under extreme conditions.
Parallel Progress and Priority Disputes
OpenAI's announcement came hours after mathematicians Tristan Buckmaster at New York University and Levent Alpöge at Anthropic released their own work. That team had used AI models including Claude and Codex to solve a related problem—the zero-viscosity Euler equations—finding a similar singularity formation scenario. Their work was published on September 7.
Buckmaster has since questioned the timing of OpenAI's efforts, alleging that the company began working on Navier-Stokes after becoming aware of his team's progress. OpenAI denies that its agents or researchers saw the independent team's work before public release, though it acknowledged it cannot rule out that de-identified usage data from the team's products may have influenced model training.
Computing Costs and Scientific Implications
The computational resources required for this breakthrough were substantial. OpenAI Chief Research Officer Mark Chen estimated the effort cost millions of dollars in compute—roughly a thousand times more than the company's previous AI mathematics experiments. Karthik Duraisamy at the University of Michigan independently estimated retail rates for similar work would exceed $6 million.
Despite resolving a Millennium Prize problem that carries a $1 million award, OpenAI has stated it does not intend to claim the prize. The company framed the release as an evaluation of its AI systems' capabilities rather than a pursuit of the monetary award.
Next Steps for the Research Community
Mathematicians and physicists will now work to verify the proof independently. The Clay Mathematics Institute has not yet weighed in on whether the solution meets the formal criteria for the Millennium Prize. Other researchers, including Anima Anandkumar at the California Institute of Technology, have also released solutions to related equations using different approaches.
The breakthrough has implications beyond pure mathematics. If verified, it may inform how engineers and scientists use the Navier-Stokes equations in applications ranging from climate modeling to biomedical research. It also represents a milestone in AI's ability to assist with high-level scientific reasoning and formal proof construction.
"This is, to me, the spectacular culmination of the arc we have seen over the last 12 months," OpenAI mathematician Sebastian Bubeck told reporters, referring to AI systems solving problems of increasing complexity.
OpenAI has published the complete proof and Lean formalization on its website and on arXiv, allowing the broader research community to examine and verify the work. The company has also made available technical documentation describing the agent architecture and computational methodology used in the discovery.
Sources: OpenAI, "On the Navier–Stokes Millennium Prize Problem," September 8, 2026. Nature, "OpenAI claims huge maths breakthrough on a famed 'Millennium Problem'," September 8, 2026. Quanta Magazine, "AI Has Solved One of Math's $1 Million Millennium Prize Problems," September 8, 2026. The Guardian, "OpenAI claims to have solved maths problem that stumped humans for decades," September 8, 2026. New York Times, "A.I. Solves a Century-Old Math Problem and Stirs a Debate Over Credit," September 8, 2026.
FIRAT Editorial Board
Institutional Research Desk · Foresight Institute of Research and Translation
The collective editorial and research translation board of FIRAT, synthesising peer-reviewed evidence, policy briefs, and division milestones across our seven foundational research pillars.
