OpenAI on Monday disclosed that an internal model it began training on August 28 has resolved more than 100 open problems "across most areas of mathematics," a claim the company made alongside the formation of a new external advisory body meant to manage the fallout from its accelerating mathematical output1,2.
The Advisory Group on Mathematics and Artificial Intelligence will be hosted at the Institute for Advanced Study in Princeton, New Jersey. Its mandate is to assess the significance of new results and coordinate their release, serving as what OpenAI called "a bridge to the mathematical community and broader public".
Scope and limits of the advisory group
Members of the group will not be paid. They can offer unsolicited advice, go public with their views, and control their own membership, giving them a degree of independence from OpenAI. The group will not, however, have leeway to slow down or redirect OpenAI's ongoing mathematical research.
ANALYSIS That structural constraint limits the body to a review-and-release function rather than a governance one, leaving OpenAI's research trajectory unchanged regardless of the group's recommendations.
The announcement follows OpenAI's earlier publication of a solution to the Navier–Stokes Millennium Prize Problem. That release drew sharp criticism from the mathematics community: 25 Fields Medal-winning mathematicians signed an open letter arguing that AI labs are threatening their intellectual work as they compete to solve famous math problems.
Mathematician backlash
The open letter from Fields Medal winners frames the dispute as one over pace and process, not capability. ANALYSIS By standing up an advisory group that can publicize dissent but cannot alter research direction, OpenAI is offering the mathematics community a consultative channel while retaining full operational control over what it trains and publishes.
OpenAI's blog post states the same model behind the Navier–Stokes result produced the additional 100-plus solutions. Training began on August 28, placing the claimed output within a roughly 24-day window before the September 21 announcement4.
The advisory group's creation arrives during a period of heightened public positioning by OpenAI. The company published a separate set of proposals on Monday calling for the United States to lead an international effort to establish global AI standards, including alignment research and recursive self-improvement control risks[1].