OpenAI released 376 mathematical results on October 6, 2026, including a claimed proof of the unique games conjecture.
The release has prompted mathematicians and research leaders to question how frontier AI systems verify advanced work and whether proprietary models remain accessible to the wider field.
The results span algebra, theoretical computer science and mathematical logic, according to the Guardian and Quanta Magazine.
The release followed OpenAI's announcement of a solution to the Navier-Stokes problem, another major mathematical challenge.
Verification Gap and Access Concerns
The Institute for Advanced Study in Princeton, New Jersey, said AI-generated mathematical arguments can create a gap between producing a result and understanding or verifying it.
"It is now the case that AI can output mathematical arguments in situations without the human who prompted it being able to understand the arguments, verify them, or take responsibility for them," the Institute said in a statement.
"We believe that human understanding of mathematics remains of paramount importance.
How, in this new era, can we work towards a new paradigm that includes human understanding of mathematics as part of responsible scholarly output?"
OpenAI said it would work with the Institute for Advanced Study to give mathematicians a voice in determining how the technology develops.
The company did not indicate that it would stop testing its models on advanced mathematical problems.
Tristan Buckmaster, a New York University mathematician who was working on the Navier-Stokes problem, questioned whether some AI results could depend on contributions supplied by researchers prompting the models.
"There's likely to be a bunch of results where they take someone's work and then take it to completion," Buckmaster said.
The advisory board also called for equitable access to AI systems used for mathematical research.
It warned that restricted access could separate AI laboratories from the broader mathematical community.
"The use of proprietary internal models by AI labs to do mathematical research risks creating a two-tier system where labs outrun the rest of the field, effectively alienating the mathematical community from its own discipline," the advisory group wrote.
The rush surrounding OpenAI's results also affected researchers working on related problems.
On September 11, 2026, Dor Minzer, a Massachusetts Institute of Technology professor, received messages suggesting OpenAI had solved the unique games conjecture.
Minzer and his graduate students Yumou Fei and Shuo Wang had recently proved a milestone result related to the conjecture.
They accelerated publication of their 95-page paper and posted it online three days after the rumors began circulating.
"The current version of the manuscript is complete mathematically, but it is not in the shape we wished to share in," the paper's authors wrote on its first page.
The unique games conjecture, first articulated by Subhash Khot in 2002, concerns the difficulty of finding approximate solutions to constraint satisfaction problems.
Its implications extend to other areas of mathematics and theoretical computer science.
Ryan O'Donnell, a theoretical computer scientist at Carnegie Mellon University, praised Minzer's work after the OpenAI announcement.
"He's put out many great papers on this topic," O'Donnell said.
O'Donnell also described the new work as a significant contribution before explaining the broader research problem addressed by the conjecture.
"A lot of work has been devoted to trying to figure out where is the boundary," O'Donnell said.
