OpenAI Establishes Independent Mathematics and AI Advisory Group
OpenAI has announced that mathematicians are establishing an independent Advisory Group on Mathematics and Artificial Intelligence, hosted at the Institute for Advanced Study. The company said a new internal model, whose training began on August 28, has resolved the Navier–Stokes Millennium Prize problem and more than 100 long-standing open problems across most areas of mathematics.
OpenAI said the rate of mathematical progress surprised its own mathematicians and prompted internal discussions about how to inform the research community so that the field can prepare and adapt. The company presented the group as an initial step in engaging mathematicians as increasingly capable AI systems produce mathematical results.
Independent advice on results and communication
The advisory group will help OpenAI assess the significance of emerging results, coordinate their dissemination, and address academic and professional standards for mathematical research. It will also advise on how AI tools might support mathematical research and learning.
The announcement follows an open letter, A Severe Misalignment of AI in Mathematics, in which mathematicians raised concerns about negative externalities created by treating solutions to open problems as a benchmark for AI systems. OpenAI said those criticisms demonstrate the need for thoughtful engagement between AI companies and the mathematics community.
The company said the group will operate independently. It may offer advice that OpenAI has not requested, comment publicly on OpenAI’s effect on mathematics, and publish its advice. Members will not be paid by OpenAI, and the group may alter its membership at its own discretion.
Scope and membership
The group will not be responsible for advising OpenAI on the pace of its internal mathematics progress. Its remit instead concentrates on reviewing and communicating results, research standards, and the potential role of the company’s tools in research and education.
Its initial members are François Charles of ENS-PSL; Camillo De Lellis and Edward Witten of the Institute for Advanced Study; Timothy Gowers of Collège de France and Cambridge; Martin Hairer of EPFL and Imperial College London; Nikhil Srivastava of Berkeley; Ulrike Tillmann of Oxford; Ravi Vakil of Stanford; and Melanie Matchett Wood of Harvard.
Why the structure matters
OpenAI described mathematics as a fundamental science whose discoveries can lead to applications well beyond the discipline. It said responsible development and deployment of mathematics-related AI therefore matter outside mathematical research, while difficult questions remain about supporting understanding and extending benefits to the wider community.
For businesses evaluating advanced AI, the announcement underlines the value of separating capability claims from reviewed findings and establishing clear processes for validation, disclosure, and responsible use when systems produce potentially high-impact research outputs.