Mathematics: Proofs, Prizes and AI-Assisted Results — 2026-09-13
The mathematical community is in a state of upheaval following OpenAI’s claim to have solved the Navier-Stokes Millennium Prize problem using 10,000 AI agents. In response, 25 Fields Medalists issued a collective warning on September 11 regarding the divergent goals between AI labs and the mathematical community, citing concerns over intellectual credit and the nature of mathematical proof.
Mathematics: Proofs, Prizes and AI-Assisted Results — 2026-09-13
Top developments
25 Fields Medalists Issue Collective Warning on AI
On September 11, 2026, a group of 25 Fields Medalists published an open letter expressing deep concern over the trajectory of AI in mathematics. The signatories argue that the objectives of AI companies and the mathematical community "diverge profoundly," particularly regarding the definition of proof and intellectual ownership. This formal stance comes just days after OpenAI announced its breakthrough on the Navier-Stokes equations, marking a significant escalation in tensions between academic mathematicians and tech giants.
OpenAI Claims Navier-Stokes Solution Amid Plagiarism Accusations
OpenAI announced on September 8 that its models solved the Navier-Stokes existence and smoothness problem, a $1 million Clay Mathematics Institute challenge, after 88 hours of computation involving 10,000 AI agents. However, the announcement was immediately shadowed by accusations from two mathematicians who claim their unpublished work may have influenced the AI’s output. The controversy has sparked intense debate over whether AI-generated proofs constitute original research or derivative synthesis of existing data.
Hodge Conjecture Rumors Spark Community Debate
Following the Navier-Stokes announcement, rumors circulated that OpenAI had also made progress on the Hodge Conjecture, another Millennium Prize problem. Terry Tao, a Fields Medalist, addressed these rumors on his blog on September 11, clarifying that while the community is excited, the situation is "messier" than viral tweets suggest. Tao emphasized that the real story lies in how the mathematical community will adapt to machine-generated results, rather than the immediate validity of specific unverified claims.
Formal Verification of Erdős Problems Continues
While the media focuses on Millennium Prizes, steady progress continues in formal verification. A recent preprint details the resolution of Erdős Problem #728 using a combination of GPT-5.2 Pro and the Aristotle formal proof assistant. The system produced a fully formalized proof in Lean 4, which has been translated into informal mathematics for peer review. This highlights the growing utility of AI in solving specific, well-defined conjectures where formal verification is feasible.
Local view
France: French media has heavily covered the backlash against AI, with Le Monde publishing the full text of the 25 Fields Medalists' warning. The article emphasizes the philosophical conflict, noting that for many mathematicians, the process of understanding is as valuable as the result itself. Pour la Science also highlighted the diversity of the recent Fields Medal winners, contrasting their human-centric narratives with the automated nature of the new AI proofs.
Germany: German outlets like Die Zeit and NZZ have focused on the "plagiarism suspicion" (Plagiatsverdacht) surrounding the OpenAI announcement. Die Zeit reported that a researcher has raised serious allegations regarding the origins of the proof, questioning the transparency of OpenAI’s training data. The coverage reflects a broader European skepticism towards proprietary AI claims in scientific discovery.
Context & numbers
- Computation Cost: OpenAI reportedly spent approximately $15 million in computational resources to solve the Navier-Stokes problem, utilizing 10,000 AI agents over 88 hours.
- Prize Value: The Navier-Stokes problem carries a $1 million prize from the Clay Mathematics Institute, one of seven Millennium Prize Problems established in 2000.
- Signatories: The warning letter against AI in mathematics was signed by 25 living Fields Medalists, representing a significant portion of the field’s most prestigious awardees.
On the radar
- Peer Review Timeline: The technical paper for OpenAI’s Navier-Stokes solution was released on September 9. Mathematicians worldwide are currently attempting to verify the proof, a process that could take months or years given the complexity of fluid dynamics.
- Hannover Math Tournament: On September 18, 2026, Leibniz University Hannover will host the "Hannoversches Mathematikturnier," a competition that may serve as a testing ground for human-AI collaborative problem-solving techniques.
- Further Formalization: Expect more preprints detailing AI-assisted resolutions of Erdős problems, as the Lean 4 ecosystem expands its library of verified theorems.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.
Powered by
