OpenAI claims AI solved Navier-Stokes problem, faces authorship dispute

OpenAI says an internal AI model, more powerful than its newly released GPT-6 Astra, found a solution to the Navier-Stokes existence and smoothness problem using 10,000 concurrent agents, with training starting August 28. Navier-Stokes is one of seven Millennium Prize Problems, each carrying a $1 million reward. Mathematician Tristan Buckmaster alleges an OpenAI researcher pressured him to drop an Anthropic-affiliated co-author from a related paper after learning of his progress through drafts he had uploaded to Codex.

Why it matters: The dispute puts OpenAI's flagship claim of independent AI mathematical discovery under scrutiny just as frontier labs race to showcase autonomous research capability — if the solution path traces back to a human mathematician's private drafts fed into Codex, it undercuts the narrative. It also raises fresh questions about how labs handle user data from coding tools when that data touches competitors' researchers.