OpenAI Faces Backlash Over Millennium Prize Solution
OpenAI's aggressive push to solve legendary mathematics problems using massive compute has triggered intense backlash from researchers who accuse the AI giant of violating academic norms.

OpenAI recently claimed a solution to the Navier-Stokes problem, one of the seven legendary Millennium Prize challenges that carry a $1 million bounty. To crack the fluid-dynamics puzzle, the company deployed roughly 10,000 AI agents and spent tens of millions of dollars in compute over just 88 hours. However, the achievement has sparked fierce controversy within the academic community. Researchers accuse the company of aggressive, competitive behavior that disregards long-standing scientific norms in its rush to beat rivals like Anthropic.
The tension peaked when NYU professor Tristan Buckmaster and Anthropic researcher Levent Alpöge were close to their own proof. Buckmaster alleges that OpenAI researcher Sébastien Bubeck offered him unlimited compute and sole paper authorship if he excluded Alpöge from the project. Buckmaster rejected the offer, calling it a bribe, and questioned whether OpenAI's Codex tool had scraped his private data. OpenAI spokesperson Laurance Fauconnet categorically denied that Buckmaster's prompts influenced the system. Meanwhile, mathematician Andreas Thom of the Technical University of Dresden raised similar concerns, noting OpenAI quietly amended a previous announcement to acknowledge his and Gábor Kun's work only after pressure.
For AI practitioners and researchers, this conflict highlights a growing trust deficit and a structural conflict of interest. As Fields Medalist Shing-Tung Yau pointed out, AI companies act as both tool providers and active competitors, raising fears that user prompts could be used to scoop academic discoveries. In response, nearly 3,900 mathematicians have signed the Leiden Declaration to promote responsible AI use. OpenAI also withdrew its sponsorship of a Caltech undergraduate hackathon following student protests. As OpenAI hints at progress on a second Millennium Prize problem, practitioners must weigh the benefits of using advanced AI assistants against the risk of losing intellectual property to the very companies hosting those tools.
This is our own summary of reporting by The Verge AI



