AIToday
Large Language ModelsAI Safety & AlignmentSemafor TechPublished: Sep 9, 2026, 10:00 JST2 min read

OpenAI solves $1 million Millennium Prize Problem

OpenAI solves $1 million Millennium Prize Problem

3 Key Points

  1. What happened

    OpenAI proved Navier-Stokes equations sometimes break down under extreme conditions, solving a $1 million Millennium Prize Problem. It deployed around 10,000 agents of an unreleased model on Aug. 28, finishing within 88 hours.

  2. Why it matters

    The question had stumped scientists for two centuries, per Scientific American. Earlier, two mathematicians rushed solutions, with one accusing OpenAI of muscling in; OpenAI denied it. This marks a shift from human-led to AI-led discovery.

  3. What to watch

    Whether the proof withstands peer review and whether OpenAI's move sparks more clashes with human researchers. The drama involves mathematician Tristan Buckmaster, who called it a "Deep Blue-Kasparov moment."

WHO IT HITSMathematicians and fluid dynamics researchers face potential competition from AI-driven discovery. Funding bodies and academic institutions may need to weigh how to credit or prioritize such machine-generated proofs.

Ask the AI about this article →

Summaries like this, in your inbox every morning.

Context & Analysis

This result came from a focused push by OpenAI, which deployed a large agentic system on a single problem. The speed—88 hours for a question open for two centuries—highlights how AI can now tackle pure mathematics, not just pattern recognition. Buckmaster's reaction suggests professional mathematicians see this as both a threat and a breakthrough.

The controversy reveals a key tension: how credit and precedence work when an AI system, guided by a company, enters a field long driven by individual insight. Buckmaster's "Deep Blue-Kasparov moment" framing suggests a parallel to when a machine first beat a human champion, but here the stakes include a $1 million prize and the future of mathematical research.

What is unresolved is whether the proof is fully accepted by the mathematical community. If it holds, it may push more scientists to rely on AI for theoretical breakthroughs, even as disputes over recognition—like the one with Buckmaster—could repeat.

FAQ
What is the Navier-Stokes problem?
It describes fluid flow, like a bathtub drain or El Niño. Scientists have not known whether the equations always work, and proving they do or don't is worth $1 million.
How did OpenAI solve it?
OpenAI used around 10,000 agents of an as-yet unreleased model on Aug. 28. The agents proved the equations sometimes break down under extreme conditions, within 88 hours.
Was there controversy?
Yes. Two mathematicians rushed out work on sub-questions, and one, Tristan Buckmaster, accused OpenAI of muscling in. OpenAI denied the claim.

Also reported by OpenAI Blog

Get the latest Large Language Models news every morning

For example, today's edition would include:

  • OpenAI's Tibo Flags Possible Pause on Pro Sign-upsITmedia AI+ · 2h ago
  • Meta's Muse agent launch could be the opposite of OpenAI's math featStratechery (Ben Thompson) · 2h ago
  • Conveo raises $50M to use AI for product insights, tackling Steve Jobs' focus group critiqueFortune AI · 2h ago

AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next articleOpenAI's ChatGPT Images 2.5 ships with better editing, speed