Has AI Solved Navier–Stokes? What We Know About the Claims and Dispute

First published September 8, 2026. Evidence reviewed through 08:21 UTC / 1:21 a.m. Pacific. This is a developing story based on public sources.

Current status: a Navier–Stokes solution has not been publicly verified by this investigation. Related fluid-equation papers and formalization materials are public. The reported internal OpenAI proof has not been available for our examination. Allegations about research conduct remain contested, and use of private drafts is unestablished. The distinctions below are essential to understanding the story.

Something significant is happening around Navier–Stokes. Mathematicians have released new work on closely related fluid equations, computer-checked proofs are part of the announcement, and Terence Tao thinks the approach could reach the famous problem. Meanwhile, social media is circulating a much stronger claim: an internal OpenAI model has already finished the job. Tao’s assessment supports excitement about a research direction; it does not establish that the reported OpenAI proof is correct. Tao’s assessment.

As of this article’s evidence cutoff, the Clay Mathematics Institute still lists Navier–Stokes among its unsolved problems. This investigation has not located a publicly available OpenAI proof resolving it. That is the defensible current status. An announcement could change the available evidence quickly, and mathematical acceptance can precede a prize decision. Clay’s unsolved list.

The story deserves attention on several grounds: the mathematics, the growing role of AI in producing proofs, and a dispute about how research involving commercial AI systems should be credited and disclosed. Those questions need separate answers.

Here is what the principal developments actually concern:

The first two rows are drawn from the released Euler, Boussinesq and IPM manuscripts. The OpenAI claim comes from Buckmaster’s statement. The last row reflects the qualifications in the separate Euler manuscript, particularly its conditional reconstruction discussion on page 13 and remaining certification work on page 67. Page counts refer to the versions downloaded for this article.

The prize problem is about what the equations guarantee. Start with an admissible smooth three-dimensional incompressible flow: can a smooth solution continue for all future time, or can it break down? A singularity is a failure of the required mathematical regularity. Computing an approximate flow for an engineering application does not answer that universal question. Fefferman’s official problem statement.

There is a crucial detail in the rules. Options A and B ask for global smooth solutions with no external force, in ordinary three-dimensional space or a periodic domain. Options C and D allow a breakdown example with a smooth external force satisfying the specified decay and periodicity conditions. Consequently, a qualifying proof for forced Navier–Stokes could resolve a version of the Millennium problem. The force must obey the full assumptions; allowing it to become singular would evade the question. Official statement, page 2.

Euler removes the viscosity term. The new Euler manuscript constructs smooth initial data and a force that remains smooth through the singular time, while vorticity—the local measure of rotation—becomes unbounded. Its geometry, regularity and forcing conditions are central to its claim. A headline that drops “Euler” or “with smooth forcing” changes the theorem being reported. Euler preprint, Theorem 1.1.

The technical advance has a history. Diego Córdoba and Luis Martínez-Zoroa had already developed a program for constructing singularities by making different spatial scales interact. Their IPM preprint was first submitted in October 2024 and revised in February 2025. The current work builds on that research. Córdoba–Martínez-Zoroa preprint.

For IPM, an important refinement is control of smoothness in both space and time. The new paper adapts the construction to a periodic domain and makes the forcing smooth in the joint variables. This distinction can disappear in a summary that calls both the earlier and newer forces simply “smooth.” The title alone is insufficient to compare the results. IPM preprint, introduction.

The Boussinesq paper gives a useful explanation of the mechanism. Begin with a background configuration that amplifies a small oscillation. The oscillation can have a small amplitude and a large gradient because its wavelength is short. Arrange for that stronger gradient to help amplify the next, even finer oscillation. Fit infinitely many stages into a finite interval, while controlling the unwanted effects introduced at each step. The hard part is making the gradients grow without losing control of the external forces or the rest of the equations. Boussinesq preprint, section 1.2.

In that construction, a relatively compact system of ordinary differential equations describes the central amplification mechanism. Turning it into a complete argument requires much more: localization, error corrections, derivative estimates and a consistent choice of scales. This helps explain why an attractive idea can still lead to a long proof. Boussinesq preprint, sections 1.2–1.3.

Expert assessment, not a completion forecast: Tao sees a plausible route to Navier–Stokes, while emphasizing enormous remaining technical difficulties. He also raises the possibility of eventually eliminating the force. These are judgments about a research direction, without a completion date or a certified Navier–Stokes theorem. His assessment does not establish how close a complete proof is. Tao’s follow-up.

The most informative way to assess closeness is to name the missing step. Here it is the successful extension to the actual viscous equation under the required assumptions, followed by scrutiny of the resulting proof. A percentage-complete estimate would hide that uncertainty. The remaining work may be amenable to computation and AI assistance, but a promising program still has to succeed.

The human dispute begins with a more compressed chronology. Buckmaster dates the Boussinesq/Euler breakthrough to August 15 and Lean verification to August 22. He calls the collaboration personal, rather than an institutional Anthropic project. His statement also mentions an unreleased hypodissipative Navier–Stokes result whose Lean verification was unfinished. Buckmaster’s statement, page 1. Hypodissipative results concern a modified dissipation model and need their own classification. Euler preprint, related work.

Reported private claim and contested account: According to Buckmaster, during September 6 calls involving Sébastien Bubeck, OpenAI described an internal, roughly 100-page proof for forced Navier–Stokes under Clay options C and D. Buckmaster reports pressure over publication and authorship, including excluding Alpöge because of his Anthropic employment, and comments he understood as career threats. He also disputes the initial characterization of the amount of human input. Buckmaster’s statement, pages 2–3.

Unknown: His account raises questions about access to their Codex drafts but does not establish that those drafts were used. His own qualifications are explicit: “I have not seen it,” and “I do not know whether our data was used.” Buckmaster’s statement, pages 3–4.

Public denial: Bubeck rejected allegations circulating about him as “false and inflammatory.” He said he had approached the discussions according to academic norms and promised further comment. His short response does not provide a detailed reconstruction of the exchanges. Both his denial and the limits of the available evidence belong beside Buckmaster’s account. Bubeck’s response.

Several questions follow, and none can be answered by counting reposts. Does a proposed proof work? How did the researchers and models arrive at it? What information did they have? What was said in the private conversations? A correct proof would answer the first question while leaving the others open. Conversely, a dispute over conduct would not establish whether the mathematics is valid.

Evidence of independent development would require more than confidence in a company or a researcher. Useful material would include a dated account of the research process, disclosed human contributions, and appropriate evidence about which inputs were available. A similar method can arise from shared published literature, information about another group’s direction, or access to unpublished material. Those possibilities should not be collapsed into a single allegation.

The posts circulating this story show how the distinctions get lost. Deane Yang’s linked post points to Buckmaster’s announcement; it does not itself announce a Navier–Stokes solution. Emad Mostaque goes further, suggesting OpenAI already has a full solution using the same approach. That suggestion exceeds what Tao’s linked exposition independently verifies. Yang’s post, Mostaque’s post.

Unconfirmed timing: Haider’s post labels an imminent OpenAI announcement as a rumor. The primary statement located here is Bubeck’s promise to say more about the allegations. A promised response does not, by itself, confirm a scheduled proof release. Relative dates such as “tomorrow” also need care: Bubeck’s post displays September 7 in Pacific time, while the same event is September 8 in UTC. Haider’s post, Bubeck’s post.

Andrew Curran’s later post points to a separate Euler result. Its authors are Adarsh Ganeshram, Valentin Duruisseaux and Anima Anandkumar. Their approach uses a physics-informed neural network to find an approximate self-similar profile for Euler on unbounded three-dimensional space, without a boundary or external forcing. The work addresses a different set of assumptions using a different computational strategy. Curran’s post, authors’ announcement.

Conditional result, with proof work remaining: The manuscript reports numerical evidence and certified ingredients, alongside a conditional nonlinear stability framework. It identifies remaining quantitative certification work; it also states that an unconditional blowup result needs an appropriate existence and continuation argument. Tao’s added commentary describes completion of the stability argument within the residual-error tolerance as a major challenge. Calling this a completed unforced Euler proof would outrun both sources. Euler manuscript, pages 13 and 67, Tao’s updated exposition.

AI’s role also requires a more precise vocabulary. In the IPM paper, the authors describe using Claude to work through prior literature and using Claude and Codex for drafting and the bookkeeping of constants and induction, under human direction. They describe their own adaptations and simplifications. This is a documented research workflow involving both mathematical judgment and model-generated work. IPM paper, discussion of language models.

The public Euler formalization’s README says Claude wrote the Lean code under Alpöge’s direction. It identifies a separate statement file and a comparator intended to check that the supplied proof establishes that statement. It also describes checks restricting the proof to the usual three foundational axioms used in classical Lean mathematics. These are concrete verification claims that other researchers can examine. This investigation inspected the documentation and theorem descriptions; it did not reproduce the large Lean build. Euler formalization documentation.

Lean matters because a proof checker can validate formal derivations under explicit assumptions. But the exact statement still matters, as do its definitions and dependencies. Lean’s documentation explains that additional axioms can undermine a proof and provides tools to audit them. For an article, “a formalization is available,” “an independent checker reproduced it,” and “experts confirmed that its statement matches the claimed theorem” describe different pieces of evidence. Lean’s documentation on axioms.

Physics-informed neural networks play another role: searching for candidate structures while penalizing violations of the governing equations. DeepMind and academic collaborators reported related work in September 2025. That history helps explain why the present moment contains several approaches, overlapping teams and different uses of the word “AI.” DeepMind’s 2025 research announcement.

The potential reward is mathematical understanding. A constructed singularity would reveal something profound about the limits of a particular idealized fluid model. It would not automatically supply a practical turbulence solver or immediate improvements in aircraft design. Those applications would require further work connecting the result and its methods to the problems engineers actually compute. Brown mathematician Javier Gómez-Serrano has emphasized both the promise of AI discovery and the substantial mathematics still needed to turn candidates into proofs. Brown’s August 26 interview.

There is also a distinction between mathematical acceptance and awarding the million dollars. Clay requires publication in a qualifying outlet, at least two years since that publication, and general acceptance by the global mathematics community before considering a proposed solution. An unchanged prize page cannot by itself refute a new proof; equally, a social-media announcement cannot substitute for the proof and its examination. Clay’s prize rules.

The next developments to watch are specific: a public Navier–Stokes manuscript with its exact assumptions; independent mathematical assessments; reproducible formal-verification results where claimed; further statements addressing the disputed exchanges; and completion or correction of the separate Euler stability work. An update should explain which of these has changed, and what remains unresolved.

For now, the strongest supported account is that an established mathematical program has produced substantial new public results with extensive AI assistance, while a reported extension to the Millennium problem awaits public examination. The importance of that progress is already clear enough to justify close attention. The final verdict belongs to the evidence still to come.

Reporting method: Kingy.ai used AI research agents to review public source material and assess potential updates. We read the released theorem statements, introductory explanations, relevant qualifications, public statements and formalization documentation. We did not independently verify every proof step, reproduce the Lean builds or numerical certificates, interview the participants, or inspect the reported private OpenAI proof. This article is reporting and analysis, rather than an independent referee report.

Update policy: We are monitoring this story every four hours for two weeks, through September 22, 2026. Agents assess new material for relevance, evidence and significance before an update is published. A released proof, substantive expert assessment, correction, direct response or evidence that changes the account can warrant an update. Repeated rumors and engagement alone do not. Material changes will be dated below; a check without new findings does not refresh the article’s evidence timestamp.

Update log — September 8, 2026: Initial publication. Separate checks of the mathematics and public dispute found no material change from the initial research. The article distinguishes public papers, unexamined private claims, conditional results, contested allegations and unanswered questions.

Featured image: AI-generated editorial illustration; not a simulation or proof visualization.