OpenAI establishes independent advisory panel for mathematics and AI collaboration

On August 28, 2026, OpenAI initiated training on a new internal model that has already resolved over 100 long-standing mathematical problems, including the Navier–Stokes Millennium Prize problem. This progress has surprised mathematicians within the company, prompting discussions on how to share these findings with the broader academic community. OpenAI emphasized the importance of responsibly deploying math-related AI capabilities, recognizing their potential for wide-ranging applications beyond mathematics. In response to concerns raised in a recent open letter by mathematicians about the risks of using open problems as benchmarks for AI systems, OpenAI has formed an independent advisory group on mathematics and artificial intelligence. The group, hosted at the Institute for Advanced Study, includes prominent mathematicians such as Timothy Gowers, Martin Hairer, and Nikhil Srivastava. The group will advise on the communication of mathematical results, the coordination of their dissemination, and the development of academic standards. It will also focus on how AI tools can support mathematical research and learning, ensuring mathematicians remain central to shaping the future of AI in the field.