OpenAI has assembled a panel of nine elite mathematicians to advise the company and other AI labs on how to engage with the mathematical research community. The advisory group, formally called the Advisory Group on Mathematics and Artificial Intelligence, or AGMAI, was announced on Monday and hosted by Princeton's Institute for Advanced Study. It arrives in the wake of a series of public missteps that left many mathematicians questioning whether OpenAI understood the discipline it had been releasing results about.

Who is on the panel and where it sits

The nine members include researchers drawn from Stanford, Harvard, Oxford, and Cambridge, with multiple Fields Medals and MacArthur grants among them. Martin Hairer, a professor of mathematics at Imperial College London and the Swiss Federal Institute of Technology in Lausanne, is one of the members. The group operates under the Institute for Advanced Study, the same institution that served as the intellectual home of Albert Einstein, John von Neumann, and J. Robert Oppenheimer. Members are not paid by OpenAI, and the group can change its own membership as it sees fit.

OpenAI stated that the group will operate independently, with the freedom to offer advice the company did not request and to make that advice public. The group will help OpenAI assess the significance of its mathematical results, advise on how to coordinate their dissemination, and counsel on academic and professional standards of mathematical research. It will also advise on how OpenAI's tools can support mathematical research and learning.

What prompted the panel

The formation of AGMAI is a direct response to a reputational crisis OpenAI created for itself through its handling of mathematical results. The company announced that its unreleased model, the same one credited with resolving the Navier-Stokes Millennium Prize problem, had also resolved more than 100 long-standing open problems across most areas of mathematics. The announcements triggered accusations of scooping researchers, failing to credit the human work underlying the models' results, and even appropriating contributions from users of OpenAI's own tools.

The controversy led many mathematicians to wonder whether OpenAI understood or even respected the community it had entered. In a blog post, Hairer addressed what he called a "torrent of misinformation" about the group's formation. He acknowledged that the group was launched after OpenAI approached some of its members, but stressed that not all members had been contacted by the company and that the organization is genuinely independent of OpenAI and other frontier AI labs. Members have not signed restrictive agreements about what they can say publicly and receive no financial compensation.

Skepticism from the community

Not everyone is convinced the panel will be effective or representative. Simon Machado, a researcher at ETH Zurich, described the group as "shrouded in mystery" and questioned how the nine members were selected. Francesco Fournier-Facio, an incoming mathematics professor at Heriot-Watt University, called the group "an ivory tower" and questioned whether it reflects the community whose work and livelihoods are being reshaped by OpenAI's push to accumulate mathematical trophies.

Kevin Buzzard, a mathematics professor at Imperial College London, went further. After months of observing what the community wanted, he wondered why OpenAI felt it needed such an elite body to understand those desires. "It wasn't proofs of hard theorems, it was better understanding of our subject," he said. "It's not entirely clear to me that you need a committee of brilliant people to hammer the point home."

Hairer himself acknowledged the limitations. The group is hardly a representative sample of the mathematical community, he said, and the small size raises questions about whether it has the breadth of expertise needed to evaluate everything coming its way. He also warned that members would "be very naïve to believe that the AI labs won't try to spin whatever we say in a way that suits their PR machine."

What the group actually does next

The group's first priority is addressing the backlog of mathematical results OpenAI claims its internal model has produced and how those results should be released. Hairer said the group is gathering input through conversations with colleagues, a public feedback form, and discussions on Proofs and Prompts, a blog that has become a central forum for mathematicians grappling with AI's impact on the field.

At this stage, the group has not settled on the advice it intends to provide. The open questions that researchers raised in interviews with The Verge remain largely unanswered: what the group will actually do, how much influence it will carry, and whether OpenAI will act on what it is told. The effectiveness of AGMAI will ultimately depend on whether it can earn the trust of a community that feels blindsided by the speed and manner of AI's arrival in mathematics.