IMPLICATOR
Twenty-five Fields Medal winners, including Terence Tao, signed a declaration posted September 11, 2026 saying AI companies’ use of mathematical problems as benchmarks is detrimental to the science and its community because the two sides’ goals are “severely misaligned.” Tao wrote that the statement grew from discussions among the medalists over the previous week and was released quickly because they considered the issue urgent. The group called it a general threat to intellectual work beyond mathematics because AI can produce results that once required people to understand a field and formulate new questions.
What Changed
- Twenty-five Fields Medal winners, including Terence Tao, signed a declaration posted September 11, 2026 saying AI companies' use of math problems as benchmarks is detrimental to the field.
- The signatories said rushed AI announcements leave too little time for writeups or citations, raising "severe attribution and plagiarism questions."
- The declaration did not call for a ban and proposed no specific rule, mechanism or enforcement.
- Critics including Lior Pachter said the signatories were right about AI labs but pointed to their own haste and the profession's record with students.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
What the declaration says
The initial signatories spanned the prize’s 1978 and 2026 cohorts, from Pierre Deligne to Yu Deng. In their account, a famous problem matters because the effort to solve it can produce methods that remain useful after the original question is settled.
The signatories described solving problems as a “tool and proxy” for conceptual understanding. A major solution has traditionally opened a long process in which mathematicians isolate new methods, connect them to prior work and teach the ideas to students. They said new ideas pass between people through discussion and careful writeups, a human transmission process that takes time.
AI labs’ rush to announce solutions can bypass that process, the declaration said, leaving too little time for a proper writeup or citations. The result raises what the group called “severe attribution and plagiarism questions.” AI-conceived ideas still require mathematicians to integrate them into the body of knowledge passed between generations.
FREE WEEKDAY MORNING BRIEFING
Follow how AI is changing research mathematics.
The Implicator Morning Briefing filters the AI news cycle to the stories worth your attention and explains their consequences. From San Francisco, every weekday at 4:45 a.m. Pacific, 7:45 a.m. Eastern.
About five minutes. No hype. No spam.
The declaration did not call for a ban. It said AI could enhance and accelerate mathematical study, then called for urgent action by mathematicians, AI companies and society. It proposed no specific rule, mechanism or enforcement.
The proof race behind the letter
OpenAI said on September 8, 2026 that an unreleased model had solved the Navier-Stokes Millennium Prize problem after 88 hours of work by roughly 10,000 agents operating in parallel. The proof has not yet been verified by the wider mathematics community. New York University professor Tristan Buckmaster and Anthropic mathematician Levent Alpöge had released related work hours earlier.
Buckmaster questioned whether OpenAI’s effort had been influenced by the approach the pair developed over several months. OpenAI denied seeing their work before it became public. The company also said it could not rule out the possibility that de-identified data from their product use had helped improve its models.
OpenAI research lead Dan Roberts said, “We recognize that the rapid progress of AI in mathematics is disruptive.” The company withdrew from sponsoring a Caltech math event on September 10, 2026 after criticism from researchers, another clash that preceded the declaration.
Know someone who'd find this useful? ✉️ Email it to a friend in one click, or they can subscribe free here.
Mathematicians challenge the signatories
Lior Pachter, reflecting in a September 12, 2026 post on his 18 years in the University of California, Berkeley mathematics department, wrote that the signatories were right about AI labs but pointed to their own haste. He also argued that the profession has failed to nurture all students with “great care.” Pachter noted that only one of the declaration’s 25 initial signatories was a woman.
Brown University doctoral candidate Kwon Hyun-woo offered another qualification. “There are many people who do not agree with the statement, but I think it is the declaration that has garnered the broadest support among those recently issued in the mathematical community,” Kwon said.
The declaration is open for more signatures. Buckmaster, the New York University mathematician, said, “The big story now in mathematics is that nobody wants to share anything.”
Frequently Asked Questions
Who signed the declaration?
Twenty-five Fields Medal winners were the initial signatories, spanning the prize's 1978 and 2026 cohorts, from Pierre Deligne to Yu Deng. Terence Tao is among them. The declaration is open for more signatures.
What does the declaration argue?
It says the goals of AI companies and the mathematical community are "severely misaligned." Solving problems is described as a "tool and proxy" for conceptual understanding, and rushed AI announcements leave too little time for writeups or citations of prior work.
Does the declaration call for banning AI in mathematics?
No. It says AI could enhance and accelerate mathematical study and calls for urgent action by mathematicians, AI companies and society. It proposes no specific rule, mechanism or enforcement.
What preceded the declaration?
On September 8, 2026, OpenAI said an unreleased model solved the Navier-Stokes Millennium Prize problem after 88 hours with roughly 10,000 agents. The proof has not yet been verified by the wider mathematics community. OpenAI also withdrew from sponsoring a Caltech math event on September 10.
Has anyone pushed back on the signatories?
Lior Pachter wrote that they were right about AI labs but pointed to their own haste and argued the profession has failed to nurture all students with "great care." Brown University doctoral candidate Kwon Hyun-woo said many people do not agree with the statement.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Marcus Schuler
Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: editor@implicator.ai
