The company says its advanced AI system has produced a potential solution to one of mathematics’ most difficult unanswered questions.
OpenAI says an artificial intelligence system has solved the Navier-Stokes problem, a mathematical challenge that has remained unresolved for decades. The problem is one of the seven Millennium Prize Problems identified by the Clay Mathematics Institute, each carrying a $1 million prize for a verified solution. The Navier-Stokes equations describe how fluids such as water and air move. Although the equations have been known for more than a century, mathematicians have struggled to determine whether smooth solutions always exist under the conditions described by the equations.
The question has implications beyond pure mathematics, as fluid dynamics is relevant to areas ranging from weather modelling and aerodynamics to engineering and physics. OpenAI’s latest announcement claims its internal AI system generated a solution suggesting that the equations can develop singularities under certain conditions. In simple terms, this would mean the mathematical description of fluid behaviour can reach a point where quantities become unbounded.
Thousands of AI agents worked on the problem.
According to OpenAI, around 10,000 AI agents worked on the problem simultaneously. The system reportedly reached its proposed solution in about 88 hours before another model verified the work.
The company has described the development as an example of increasingly capable AI systems moving beyond conventional question-answering and towards research tasks. OpenAI has previously reported AI-generated advances in mathematics, including work on long-standing problems in discrete geometry.
Mathematical verification remains crucial.
Despite the significance of the announcement, an AI-generated solution isn’t automatically accepted as a mathematical proof. Independent mathematicians must examine the argument, identify any gaps and determine whether the result meets established standards of proof. This distinction matters for the Navier-Stokes problem because of its complexity and the long history of unsuccessful attempts to resolve it.
OpenAI says it is releasing the proposed solution along with a formal proof written in Lean, a computer-assisted theorem-proving system. The formalisation is intended to make the argument easier to check computationally.
If independent researchers validate the Navier-Stokes result, it could represent a significant moment for both mathematics and artificial intelligence. For now, however, the claim remains subject to the scrutiny that has always been central to mathematical discovery: a result must withstand rigorous independent verification before it can be considered solved.
