OpenAI has announced a solution to a Millennium Prize problem, one of mathematics' most prestigious open challenges, but the achievement has sparked unease rather than celebration among many mathematicians. Critics argue OpenAI is approaching the field less as a collaborative contributor and more as a well-funded competitor racing to claim prizes, with little regard for academic norms or the researchers whose years of work it may be building on. The controversy centres on accusations from NYU professor Tristan Buckmaster, who claims OpenAI has not adequately explained whether his own work, done using its Codex tool, contributed to its breakthrough on the Navier-Stokes problem.

OpenAI says it used roughly 10,000 AI agents and tens of millions of dollars in compute to solve the Navier-Stokes problem, concerning fluid flow, in just 88 hours, after learning that Buckmaster and Anthropic researcher Levent Alpöge were independently close to a solution. Buckmaster alleges that OpenAI researcher Sébastien Bubeck offered him near-unlimited compute and sole authorship on the announcement paper in exchange for excluding Alpöge from credit, an offer he says he rejected as a "bribe". OpenAI has denied that Buckmaster's prompts or data influenced its model's training, though it has previously stopped short of ruling out that his broader product usage played a role, a denial he says he views with scepticism.