A group of 25 Fields Medallists — every initial signatory a winner of mathematics' highest honour — has published a formal declaration warning that the race to showcase artificial intelligence on mathematical benchmarks is damaging both the science of mathematics and the community that sustains it. The statement, titled "A Severe Misalignment of AI in Mathematics," went live on Friday, and among its initial signatories is Terence Tao, one of the most influential mathematicians of his generation.
The declaration is published at mathandai.org, and within a day it had gathered more than 1,300 additional endorsers from universities and research institutes around the world, according to the site's public signatory list. Tao confirmed his participation in a post on his personal blog, writing that he was "proud to be among the list of 25 initial signatories — all Fields Medallists." For anyone tracking the latest AI developments, the intervention marks one of the most pointed critiques yet from within elite mathematics about how AI labs measure — and publicize — their models' capabilities. It dominated discussion on Hacker News on Friday, drawing hundreds of comments.
What the Declaration Says
The signatories do not dispute that AI systems have become remarkably capable. "Over the last few months, the mathematical capabilities of LLMs have improved dramatically, to the point that they can solve major outstanding problems in many fields of mathematics," the declaration states. The objection is to what happens next: AI companies using those solutions as competitive benchmarks, which the authors argue is "detrimental to the science of mathematics, and to the mathematical community."
The core claim is stark: "The goals of the AI companies and the goals of the mathematical community are severely misaligned."
The declaration argues that solving famous problems has traditionally served as a sign of new insight — the beginning of a long communal process of talks, discussion, simplification, and eventual textbook write-ups — rather than an end in itself. "Solving problems is only a tool and proxy for achieving the primary goal of conceptual understanding and insight," the authors write. "Forgetting this in the world of AI may turn the tool against the primary goal."
The most quoted line captures the worry: "the mass production at faster and faster pace of 'true/false' statements could destroy fertile ground instead of breathing life into new ideas."
Attribution, Plagiarism, and Rushed Announcements
The signatories also raise a practical grievance. "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," the declaration states. "As in all creative professions, this raises severe attribution and plagiarism questions."
The authors warn of a second-order effect: without "willing mathematicians who must take care of their development and integration into the mathematical canon," ideas conceived by AI would never become fully developed, and "the crucial human transmission chain between mathematicians would be lost."
An Unusual Show of Urgency
The process itself was unusual. Tao acknowledged on his blog that the statement grew out of discussions among the signatories "over the last week," and apologized for not running a more consultative process. "We decided that the urgency of the situation was such that we needed to release a statement sooner rather than later," he wrote.
The declaration follows the model of the earlier Leiden declaration, and the organizers explicitly invite further signatures. Tao also pointed readers to recent coverage of the statement by The Economist. The initial 25 signatories were all Fields Medallists, a deliberate signal — mathematics' most decorated researchers speaking with one voice. The broader endorsement list already includes mathematicians from MIT, ETH Zurich, the Max Planck Institute, Northwestern, NYU, and dozens of other institutions.
A Warning Beyond Mathematics
While the statement is rooted in mathematics, the authors frame it as a template for a much wider problem. "We are witnessing a general threat to intellectual work, with misalignment between the outcome of the use of AI and its initial purpose," the declaration reads.
The argument will resonate far beyond mathematics departments. In many professions, the authors note, years of training have traditionally served not only to produce final answers but to develop understanding and the ability to formulate new questions. As AI systems become capable of producing the results of such work directly, "these goals cease to align." The declaration describes the issues facing mathematicians as "similar to issues that other scientific and creative professions are facing, and indicate issues that all of humanity might face."
Not a Rejection of AI
Notably, the statement is not a call to halt AI development. "AI offers the potential of enhancing and accelerating genuine mathematical study and understanding," the authors write, while conceding that "mathematics as a profession will need to adapt."
The intervention lands in a charged moment. Claims that AI systems have solved or made major progress on celebrated open problems have repeatedly made headlines, and earlier this year a separate dispute over an OpenAI claim related to the Navier–Stokes problem showed how contentious such announcements can become. The Fields Medallists' declaration is effectively the mathematics community's most senior response to that dynamic: a request that AI labs stop treating centuries-old open problems as leaderboard entries.
Whether AI companies respond — by consulting mathematicians before announcements, crediting prior work, or rethinking benchmarks — remains to be seen. What is already clear is that the people who know these problems best are no longer watching from the sidelines.
