OpenAI announced that an unreleased internal model solved the Navier-Stokes equation, a complex mathematical question carrying a $1mn prize from the Clay Mathematics Institute. However, New York University mathematician Tristan Buckmaster, who had been working with Levent Alpöge of rival Anthropic, has raised claims of foul play. Chief executive Sam Altman defended OpenAI, saying the lab faced “unfounded accusations of plagiarism.” The dispute highlights the intense rivalry between OpenAI and Anthropic spilling into academic controversies.

Buckmaster argued OpenAI raced to beat him by using existing approaches and throwing significant computing resources at the problem. He and Alpöge announced three proofs related to the Euler equation, using tools from Google Deep Mind, Anthropic, and OpenAI. Buckmaster claimed conversations with OpenAI researcher Sébastien Bubeck became acrimonious, while Bubeck denies the characterisation.

Altman said OpenAI would have “greatly preferred coordination” and offered to let the other team publish first.