Twenty-five Fields medallists declare a misalignment of goals
Mathematicians published a declaration: the race by companies to solve problems as a benchmark harms mathematics, because results are announced in a rush, without proper writeup or citation of predecessors.
Why it matters
The most decorated people in the field stated jointly, with signatures and a date, that the way results are being published destroys the chain of exposition and checking their discipline rests on.
The declaration, titled A Severe Misalignment of AI in Mathematics, was published on 11 September 2026 at mathandai.org with DOI 10.5281/zenodo.22737750. It carries 25 initial signatories, all Fields Medallists, spanning cohorts from 1978 to 2026; among them Artur Avila, Manjul Bhargava, Pierre Deligne, Hugo Duminil-Copin, James Maynard, Peter Scholze and Terence Tao, who posted the text on his own blog the same day. A French version appeared in Le Monde. The text states: "But solving problems is only a tool and proxy for achieving the primary goal of conceptual understanding and insight. Forgetting this in the world of AI may turn the tool against the primary goal. Indeed, the mass production at faster and faster pace of 'true/false' statements could destroy fertile ground instead of breathing life into new ideas. Often these solutions are announced in a rush, leaving no time for a proper writeup, the isolation of new methods and ideas, and citing relevant previous work of others. As in all creative professions, this raises severe attribution and plagiarism questions." The declaration followed OpenAI's publication of ten results on 1 August, its Navier-Stokes claim of 8 September, and Tristan Buckmaster's accusation the same day that the company had pressured him over crediting an Anthropic collaborator.