OpenAI announced on September 8, 2026, that an unreleased artificial intelligence model solved the Navier-Stokes existence and smoothness problem. The Navier-Stokes problem is one of mathematics’ seven Millennium Prize Problems, utilizing 10,000 AI agents working in parallel over 88 hours to produce the breakthrough.
The announcement represents one of the most significant demonstrations yet of artificial intelligence tackling advanced mathematical research. The Navier-Stokes equations describe the motion of fluids such as water and air, asking whether solutions remain well-behaved or develop singularities under certain conditions. OpenAI’s proof claimed to show such a breakdown, answering a problem that had remained unresolved for decades.
Compute Scale and the Mechanics of the Breakthrough
To tackle the fluid motion puzzle, OpenAI deployed unprecedented computing resources on a single task. According to Nature, the model first answered a simplified version of the question in 50 hours using 1,000 AI agents. Researchers then increased the computational power for the full Navier-Stokes problem.
The unreleased model utilized up to 10,000 AI agents working in parallel over the course of 88 hours.
OpenAI released a paper detailing its proposed solution, including a proof written in Lean, a programming language and theorem-proving system used to verify mathematical arguments formally. OpenAI computer scientist Ven Chandrasekaran explained the implications of the proof during a press briefing.
Because this behavior is physically impossible for a real fluid, it suggests that under certain circumstances, the equations may not be a reliable mirror of physical reality.
Origins of the Millennium Prize Problems and Human Parallel Research
The Navier-Stokes problem was one of seven Millennium Prize Problems designated by the Clay Mathematics Institute in the year 2000. Founded by American businessperson Landon T. Clay, the institute offered a $1 million prize for the first correct solution to each problem. Before OpenAI’s announcement, only one of the seven problems had been solved.
OpenAI acknowledged that it intensified its work on the Navier-Stokes problem on September 1, 2026, after learning that other mathematicians were pursuing a similar direction. Mathematicians Levent Alpöge of Harvard University and Tristan Buckmaster, a professor of mathematics at New York University, had been exploring similar research.
On September 7, Alpöge and Buckmaster released a paper detailing a solution for the fluid equations in the simplified case where the fluid has no viscosity, using models from rival AI company Anthropic alongside OpenAI’s Codex and Astra models. OpenAI stated that it did not see any of their work through any means until it was released publicly.
Community Response and Mathematical Scrutiny
The announcement has sparked mixed reactions across the mathematical community, balancing excitement over technological milestones with concerns about human understanding. Terence Tao, a professor at UCLA, described the questions as lighthouses
that serve as focus points for human scientists.
At the same time, Tao has publicly warned that advanced AI systems could end up damaging the field if machines solve difficult problems without human input. He compared the process to lifting weights at the gym, noting that AI can solve questions without gaining the instructional value that humans experience through struggle.
OpenAI researchers noted that human ideas still played a foundational part in guiding the investigation. Dan Roberts described the internal research team’s role as functioning like a bumble bee cross-pollinating across different groups and delivering different bits of information.
The proposed proof will require rigorous scrutiny by the wider mathematical community before the problem can be considered definitively settled.
