OpenAI’s claim, mathematicians push back

What's happened

OpenAI has published a proposed solution to the Navier‑Stokes Millennium Prize problem after running roughly 10,000 internal AI agents for about 88 hours. Two mathematicians, Tristan Buckmaster and Levent Alpöge, have said OpenAI accelerated work after learning of their AI‑assisted progress and raised questions about whether de‑identified product data influenced the models.

What's behind the headline?

What happened

OpenAI has run roughly 10,000 coordinating AI agents on the Navier‑Stokes existence and smoothness question and has published a proposed resolution after about 88 hours of agent work and a further verification pass. The company has said its internal model is more capable than its public GPT‑6 Astra and that the effort cost millions in compute.

The central friction

• Two mathematicians, Tristan Buckmaster and Levent Alpöge, have said they were working on related results and that OpenAI accelerated its effort after hearing of their progress. Buckmaster has published a timeline alleging OpenAI approached him and discussed coordinating announcements. OpenAI denies accessing their unpublished work but admits it "cannot rule out" that de‑identified product data helped improve its models.

Why this matters

• Credit and data governance are now active battlegrounds in science: researchers will reduce willingness to test frontier ideas in proprietary models if they believe companies can front‑run or absorb their work.
• The economics of compute gives large labs a decisive advantage: a problem that took a human community years can be resolved in days with millions of dollars of compute and thousands of agents.

Likely outcomes

• The Clay Mathematics Institute will not award the Millennium Prize until peer review and a two‑year acceptance period proceed; that process will be deliberate and public scrutiny will intensify.
• Academic norms around attribution, data access and private‑model use will tighten. Universities and funders will push for clearer rules about what material users can expect to remain private when they run research through commercial AI tools.
• Math departments will shift roles: more effort will go to verifying and interpreting machine‑produced proofs, and graduate training will emphasise conceptual understanding and verification skills.

Bottom line

This episode will force a realignment between large AI labs and the research community. OpenAI's technical claim will be judged on mathematical grounds, but the debate over credit and data will shape how mathematicians deploy AI going forward.

How we got here

Navier‑Stokes is one of seven Millennium Prize Problems set by the Clay Mathematics Institute. OpenAI has applied a new, more capable internal model and coordinated agents to tackle hard math problems, and it has said it will not claim the $1m prize while the mathematical community evaluates any proof.

Our analysis

The New York Times Business frames the dispute as a cultural fight inside mathematics, quoting Fields medalists who argue that solving famous problems without human understanding "could destroy the entire mathematical ecosystem" and referencing Terence Tao's warning of "strip‑mining" mathematics. (New York Times Business) The Guardian has emphasised the sense of betrayal among mathematicians and the attribution issue: it reports that many feel OpenAI "didn’t give sufficient credit," and quotes Tristan Buckmaster on concerns that his AI‑assisted work was seen by OpenAI researchers. The Guardian stresses that mathematicians still value human understanding over flashy solutions. (The Guardian, multiple pieces) Business Insider and CNBC focus on the timeline and cost. Business Insider notes the roughly 10,000 agents and an estimated compute bill in the millions; CNBC details OpenAI's account that agents exchanged millions of messages and that the company began its concentrated effort after hearing rumours. Both cite Buckmaster’s public timeline and OpenAI's statement that it "did not see any of their work" while acknowledging it "cannot rule out" that de‑identified data informed model improvements. (Business Insider UK; CNBC) TechCrunch, France 24 and BBC provide colour on the recruitment and contact between OpenAI staff and external academics. TechCrunch reproduces Buckmaster’s claim that OpenAI proposed publication arrangements and that Sebastian Bubeck asked for author changes; it also quotes Bubeck’s denial that he asked for removal of credit. France 24 and the BBC emphasise the compute scale and the Clay Institute's intentionally slow, rigorous prize evaluation. (TechCrunch; France 24; BBC) Across outlets, two threads recur: the technical claim (OpenAI’s agents reached a proposed proof) and the procedural dispute (whether OpenAI used privileged information or whether its outreach and offers breached academic norms). Direct quotes underline the tension: Buckmaster says he "does not know whethe

Go deeper

• How will the Clay Mathematics Institute handle a machine‑produced proof when peer review begins?
• Will universities restrict researchers from running unpublished work through commercial AI products?
• What rules will labs set to prevent using researchers' private prompts or sessions in frontier research?

More on these topics

• OpenAI - Artificial intelligence company

  OpenAI is an artificial intelligence research laboratory consisting of the for-profit corporation OpenAI LP and its parent company, the non-profit OpenAI Inc.

• Anthropic - Artificial intelligence company

  Anthropic PBC is a U.S.-based artificial intelligence startup public-benefit company, founded in 2021. It researches and develops AI to "study their safety properties at the technological frontier" and use this research to deploy safe, reliable models for

• United Kingdom - Country in Europe

  The United Kingdom of Great Britain and Northern Ireland, commonly known as the United Kingdom or Britain, is a sovereign country located off the north­western coast of the European mainland.

• New York University - Private university in New York City, New York

  New York University is a private research university based in New York City. Founded in 1831 by Albert Gallatin as an institution to "admit based upon merit rather than birthright or social class", NYU's historical campus is in Greenwich Village.

• The Guardian - Newspaper

  The Guardian is a British daily newspaper. It was founded in 1821 as The Manchester Guardian, and changed its name in 1959. Along with its sister papers The Observer and The Guardian Weekly, The Guardian is part of the Guardian Media Group, owned by the S

• Fields Medal - Award

  The Fields Medal is a prize awarded to two, three, or four mathematicians under 40 years of age at the International Congress of the International Mathematical Union, a meeting that takes place every four years.

• United States - Country in North America

  The United States of America, commonly known as the United States or America, is a country mostly located in central North America, between Canada and Mexico.

• Terry Tao - Australian-American mathematician

  Terence Chi-Shen Tao FAA FRS is an Australian-American mathematician. He is a professor of mathematics at the University of California, Los Angeles, where he holds the James and Carol Collins chair.

• New York City - US State

  New York is a state in the Northeastern United States. New York was one of the original thirteen colonies that formed the United States. With more than 19 million residents in 2019, it is the fourth-most-populous state.

• Codex - Ancient manuscript generally taking the form of the modern book: bound leaves of paper or parchment with information inscribed on both faces, attached at a center fold. Precursor of the modern book.

  The codex was the historical ancestor of the modern book. Instead of being composed of sheets of paper, it used sheets of vellum, papyrus, or other materials. The term codex is often used for ancient manuscript books, with handwritten contents.