OpenAI accused of using mathematician’s private research for unverified Navier-Stokes proof
Mathematician Tristan Buckmaster alleges OpenAI used his and colleague Levent Alpöge’s proprietary research approach to produce an unverified 100-page proof of finite time blow-up for the forced Navier-Stokes equations, a Millennium Prize Problem. Buckmaster claims OpenAI proposed excluding Alpöge from authorship due to his employment at Anthropic. OpenAI’s Sébastien Bubeck denies the allegations as “false and inflammatory.” The dispute highlights tensions over credit, data privacy, and competitive dynamics in AI-assisted research.
Editorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page itself is projected from evidence records.
- Current automated evidence projection
Cross-source coverage
Common ground
- Both sides agree that a 100-page proof no one outside OpenAI can verify is not a real scientific contribution.
- Both agree that OpenAI's secrecy is a major problem that undermines trust in their claimed breakthrough.
- Both agree that the Millennium Prize should demand the proof be released for independent review.
- Both agree that the current system lacks a mechanism to verify AI-generated proofs at scale.
Points of contention
- One side sees the timing of OpenAI's result as suspicious evidence of corporate theft, while the other sees it as a plausible outcome of AI acceleration.
- One side argues that OpenAI's pattern of behavior and vague terms of service shift the burden of proof onto them, while the other insists the burden stays on the accuser without direct evidence.
- One side views the data ethics question and the verification crisis as two sides of the same power problem, while the other sees them as separate issues and focuses on verification as the concrete fix.
Blind spots
- Both sides overlook the possibility that OpenAI's multi-agent system could have independently replicated the result using only published literature, without any data theft.
- Neither side fully addresses how academic norms and credit systems need to evolve when AI tools are owned by private companies.
- The debate ignores the role of the broader research community in pressuring OpenAI for transparency, rather than relying solely on the Millennium Prize.
WorldAttention’s read
This debate boils down to a clash between a verification crisis and a power crisis. Both sides agree that OpenAI's secret 100-page proof is not science and must be released for independent review. The main disagreement is whether the timing and data concerns are evidence of corporate theft or just a coincidence explained by AI acceleration. The neutral side argues that without direct evidence, the data theft accusation is speculation that distracts from the fixable verification problem. The western side argues that OpenAI's control over tools, data, and narrative creates a power imbalance that makes secrecy itself the scandal. The real blind spot is that neither side fully explores how academic credit and peer review must be redesigned for an era where private companies own the infrastructure of discovery. The path forward is clear: demand transparency from OpenAI, release the proof, and build new verification systems for AI-generated research. Until then, this remains a press release with equations, not a breakthrough.
Wire timeline
OpenAI Claims to Solve Millennium Navier-Stokes Problem, Sparking Academic Dispute
OpenAI announced that an internal AI model more powerful than GPT-6 Astra solved the Navier-Stokes equations, one of the seven Millennium Prize Problems, using 10,000 agents in 88 hours. The claim immediately triggered a major academic controversy. NYU mathematician Tristan Buckmaster published a statement alleging that OpenAI's team, led by Sébastien Bubeck, attempted to pressure him into a publication agreement that would have excluded his collaborator Levent Alpöge (an Anthropic employee) from authorship. Buckmaster also raised concerns that OpenAI may have used his private research data stored on its Codex platform to train the model. OpenAI's blog acknowledged the possibility of using de-identified user data. The dispute escalated on social media, with Bubeck denying the allegations and releasing text messages he claimed showed good faith. Mathematician Terence Tao had predicted days earlier that an opaque AI solution to this problem could be a net negative for mathematics. The solution reportedly proves finite-time blowup with a smooth external force, a path also being pursued by Buckmaster and Alpöge.
OpenAI claims Navier-Stokes breakthrough amid cheating allegations from mathematician
OpenAI announced that a multi-agent AI system solved the Navier-Stokes Millennium Prize Problem, proving conditions under which the equations 'blow up.' However, mathematician Tristan Buckmaster of NYU's Courant Institute alleges OpenAI used his and colleague Levent Alpöge's approach without credit. Buckmaster says he and Alpöge used AI models from both Anthropic and OpenAI to develop an almost identical solution over nearly a year, with rapid progress from mid-August. He claims OpenAI researchers admitted on a September 6 call they only attempted the problem in the past week after rumors Anthropic would announce a solution. The controversy raises questions about credit, data use, and the role of AI in mathematical research. The article also covers Google DeepMind's AI predicting genetic mutation impacts, OpenAI agents swarming a German wiki, Mistral's $24.4 billion valuation, and American pessimism about AI.
OpenAI CEO says team acted with integrity, offered rival team prize for math problem solution
In a post on X, Sam Altman, CEO of OpenAI, described a weekend of discussions with his team regarding a math problem solution. He stated that the team acted with integrity and generosity throughout. Initially, they believed another team had also solved the problem and wanted a joint release. Upon learning the other team had Euler but not Navier-Stokes, OpenAI offered to let them go first, suggested they receive the prize, and proposed Tristan as lead author on a rewrite of the OpenAI proof. Altman noted difficulty in offering the same to Levent, an Anthropic employee who was unwilling to communicate. He said OpenAI was open to other solutions and preferred coordination, but did not rush to publish despite the other team's lack of communication. The team was threatened with unfounded accusations of plagiarism. Altman noted that the approaches now appear different and that OpenAI's latest model can solve many other math problems. He confirmed the effort was prompted by rumors that Anthropic's models had solved a millennium problem.
Show 4 older updatesHide older updates
NYU mathematician accuses OpenAI of unfair competition in Millennium Prize contest; Bubeck denies
NYU mathematics professor Tristan Buckmaster and Anthropic mathematician Levent Alpöge have announced preliminary results for three proofs addressing the Millennium Prize Problem concerning the existence and smoothness of solutions to the Navier-Stokes equations. They claim that information about their research progress was leaked to OpenAI, which subsequently used substantial computational resources to pursue the proof along their unique approach. The dispute highlights trust issues arising from AI involvement in mathematical research, with OpenAI's Sebastien Bubeck denying the allegations. The original TechCrunch article details the conflict and public responses from both parties.
Tristan Buckmaster Accuses OpenAI of Using His Private Codex Data for Unverified Navier-Stokes Breakthrough
Tristan Buckmaster has reportedly accused OpenAI of developing an unverified breakthrough in the Navier-Stokes equations, a major unsolved problem in physics and mathematics. According to the report from TC, Buckmaster alleges that OpenAI used data and approaches extracted from his private Codex research sessions without authorization. The accusation centers on the claim that OpenAI leveraged proprietary insights from Buckmaster's work to achieve the purported breakthrough, which has not been independently verified. This raises significant concerns about intellectual property rights and research ethics in the AI industry. The Navier-Stokes equations describe fluid motion and are notoriously difficult to solve, making any claimed breakthrough highly significant. The incident highlights ongoing tensions between individual researchers and large AI companies over data usage and credit. OpenAI has not yet publicly responded to the allegations. The story was first reported by TechCrunch (TC).
Researcher Tristan Buckmaster says OpenAI model produced 100-page Navier-Stokes proof
Tristan Buckmaster, a mathematician, claims that OpenAI informed him that one of its internal AI models generated an approximately 100-page proof of finite time blow-up for the forced Navier-Stokes equations. If verified, this would be a landmark achievement in mathematics. However, Buckmaster states he has not personally seen or verified the proof. The post also details a controversy: Buckmaster alleges that OpenAI researcher Sébastien Bubeck proposed publication arrangements that would exclude Levent Alpöge from authorship, partly because Alpöge works at Anthropic. Buckmaster further claims that after he threatened to disclose this situation, he was asked, "Why would you ruin your career?" He also asserts that OpenAI's effort began only after information about his own research reached the company. The post suggests the situation is likely to escalate.
OpenAI AI allegedly solved Navier-Stokes problem, sparking credit dispute, Buckmaster says
A heated debate has erupted over claims that OpenAI's AI solved the Navier-Stokes Millennium Prize Problem. Mathematician Tristan Buckmaster says OpenAI privately told him on September 6 that an internal model had produced a proof for forced Navier-Stokes, potentially taking the final step toward solving the problem. Buckmaster and colleague Alpöge had previously used Claude and OpenAI models to establish how smooth fluid flows can develop singularities, advancing research relevant to the problem but not solving it. Buckmaster disputes OpenAI's description of minimal human involvement, reporting that discussions revealed a team effort with multiple attempts and substantial compute. He also claims publication proposals excluded Alpöge as coauthor due to his employment at Anthropic. OpenAI's Sébastien Bubeck has publicly rejected the allegations as 'false and inflammatory.' The claim remains unverified, and Buckmaster has not seen the proof. The controversy highlights tensions over scientific credit in AI-assisted research.