On September 11, Terence Tao, Yu Deng, Peter Scholze and 25 other Fields Medalists published, as initial signatories, a declaration titled "A Severe Misalignment of AI in Mathematics." The Fields Medal is often called the Nobel Prize of mathematics, and a joint statement from 25 living medalists is extraordinarily rare in the history of the field.
The declaration does not deny AI's progress; it openly acknowledges a change already under way: over the last few months, the mathematical capabilities of large language models have improved dramatically, to the point where they can solve major outstanding problems in many fields of mathematics. But that is precisely where the problem begins — "the goals of the AI companies and the goals of the mathematical community are severely misaligned." The push by AI companies to solve mathematical problems as a benchmark, the signatories argue, is detrimental to the science of mathematics and to the mathematical community.
In the view of these top mathematicians, a famous open problem matters not simply because a final "right" or "wrong" answer is produced. Over the long term, significant unsolved problems act more like landmarks and lighthouses of the mathematical world: solving one usually means researchers have discovered new insights, methods, or concepts. Those results then need to be reported, discussed, and simplified before they can become textbook material for graduate students — or even undergraduates — and some ideas take decades or centuries to become widely used tools. The AI industry's "benchmark grinding" turns problem-solving into a muscle-flexing KPI: rushed answers, missing attribution, citation and review rigor — and it raises severe attribution and plagiarism questions. The declaration frames this as "a general threat to intellectual work," part of broader alignment issues affecting other scientific and creative professions and society as a whole.
The signatories span Fields Medal years from 1978 to 2026, including Pierre Deligne, Maxim Kontsevich, Maryna Viazovska, James Maynard, and 2026 medalist Yu Deng. Terence Tao posted the declaration on his blog, noting that the text grew out of roughly a week of discussion; as of September 12 it had more than 1,694 endorsements.
This is not the first time mathematicians have voiced concern. At the International Congress of Mathematicians in Philadelphia in late July, Tao delivered a lecture titled "Mathematics in the Age of AI," comparing the moment to the foundational crisis of the early 20th century. This time, after AI cracked the Navier-Stokes equation in 88 hours and casually brushed aside a million-dollar prize, mathematicians decided to stop watching from the sidelines. AI can solve problems — but the soul of mathematics was never in the answers.