OpenAI disclosed on September 21, 2026, that a new internal model trained beginning August 28 has resolved the Navier-Stokes Millennium Prize problem and more than 100 other long-standing open problems across most areas of mathematics. The company is now working to establish channels with the mathematics community, acknowledging that the pace of progress has surprised even its own researchers and requires careful coordination before broader deployment.
The model's rapid advance prompted internal discussions about how to inform and prepare the field. OpenAI described mathematics as a fundamental science where novel discoveries carry wide-ranging applications, making responsible development important beyond the discipline itself. The company said it is "working through how broader deployment of math-related AI capabilities can be done responsibly."
An independent advisory group
OpenAI is partnering with mathematicians who have formed an independent advisory group, hosted at the Institute for Advanced Study. The group will operate with full independence from the company. Its members will not be paid by OpenAI, and the group can change its membership as it sees fit. It also has the freedom to offer unrequested advice, comment publicly on OpenAI's impact on mathematics, and publish its recommendations.
The group's remit covers reviewing emerging results, assessing their significance, and advising on how to coordinate dissemination. It will also advise on academic and professional standards of mathematical research and on how OpenAI's tools can support research and learning. One boundary is explicit: the group will not advise on pacing OpenAI's internal progress in mathematics.
Mathematicians' concerns
The move follows an open letter titled "A Severe Misalignment of AI in Mathematics," in which mathematicians raised concerns about the negative externalities of using open problem-solving as a benchmark for AI systems. OpenAI cited the letter directly, saying the criticisms "highlight the need for thoughtful engagement of AI companies with the math community." The advisory group is intended as a bridge to both the mathematical community and the broader public.
"We want to put capable tools in mathematicians' hands so they can pursue the questions they know best and develop new ideas," the company said. It described working with the group as a first step toward addressing difficult questions about how AI can support mathematical understanding and how benefits can reach the wider community.
Who is on the group
The initial nine members span institutions across Europe and the United States:
- François Charles (ENS-PSL)
- Camillo De Lellis (IAS, GSSI)
- Timothy Gowers (Collège de France, Cambridge)
- Martin Hairer (EPFL, Imperial College London)
- Nikhil Srivastava (Berkeley, Simons Institute)
- Ulrike Tillmann (Oxford, INI)
- Ravi Vakil (Stanford)
- Edward Witten (IAS)
- Melanie Matchett Wood (Harvard)
Why this matters for science and research professionals
A single model resolving over 100 open problems signals a shift in how mathematical research gets done. For researchers, the immediate question is not whether AI can contribute to mathematics but how to integrate tools that operate at this speed and scale into existing workflows. The advisory group's work on dissemination standards and tool support will shape what that integration looks like. Professionals working at the intersection of AI and research may want to follow the group's public advice as it emerges, since it will likely set precedents for how AI companies engage with other scientific fields. Those looking to build skills in applying AI to research problems can explore AI for Scientists Courses and AI Research Courses to stay current with practical techniques.
Your membership also unlocks:
