± HEALTH In our modern society, digital devices, social media platforms, and artificial intelligence (AI) tools have b…+ CLIMATE The growing complexity of electric vehicle charging station (EVCS) operations—driven by grid constraints, ren…± HEALTH Objectives This study aim to develop, compare and internally validate machine-learning models for predicting…+ HEALTH Diabetes management faces a number of obstacles, such as fragmented healthcare data, privacy concerns, poor e…− SCIENCE Background Craniofacial soft tissue thickness (CFSTT) is a critical parameter in forensic facial reconstructi…+ HEALTH Background Technological innovation in total joint arthroplasty (TJA) has largely focused on intraoperative p…± HEALTH Infective endocarditis (IE) continues to be an often fatal condition despite improvements in cardiac surgical…± HEALTH Purpose of review Systemic sclerosis-associated interstitial lung disease (SSc-ILD) is the leading cause of d…
TruaceTracing the truth around AIFriday, September 11, 2026
Subscribe
Science·The Trace·Dual reading·Published 2026-09-09

AI system tackling the Navier-Stokes Millennium Prize Problem

Source article: OpenAI claims to have solved maths problem that stumped humans for decades

Abstract: OpenAI claims to have solved a major mathematics problem that has stumped humans for nearly a century after spending millions of dollars on the artificial intelligence-led endeavour. The company behind ChatGPT said it had cracked the Navier-Stokes problem, one of seven Millennium Prize Problems published by the Clay Mathematics Institute to highlight some of the biggest unsolved puzzles in the field. However, the announcement swiftly became mired in controversy after mathematician Tristan Buckmaster, a professor…

TRV-2026-1020JournalismPermanent record — cite & verify
Trace impact reading

Contested: both sides are scored from claims and sources, not community votes.

P 57The P score combines the specificity and measured human impact of the grounded problem claim with the strength of this Trace’s cited sources.G 60The G score combines the specificity and measured human impact of the grounded gain claim with the strength of this Trace’s cited sources.
OpenAI claims to have solved maths problem that stumped humans for decades
The quick read

On 8 September 2026, OpenAI claimed its internal AI system, described as more powerful than GPT-6 Astra, solved the Navier-Stokes Millennium Prize Problem using about 10,000 autonomous agents working for 88 hours, with GPT-6 Astra taking about 17 hours to verify the proof that the fluid equations can blow up.

The claim matters because it would mark AI encroaching on elite mathematics previously stumped for nearly a century, but it remains uncertain due to disputed provenance, with questions about whether work stored in Codex was visible and whether independent mathematicians and the Clay Institute will accept the proof.

Main points
  • OpenAI said it tackled the Navier-Stokes problem with an internal system more powerful than its latest GPT-6 Astra model using about 10,000 AI agents.
  • The company said the agents reached a solution in 88 hours and GPT-6 Astra verified it in about 17 hours, with a proof suggesting the equations occasionally blow up.
  • Mathematician Tristan Buckmaster said his in-progress work with another researcher was stored in OpenAI's Codex model, potentially making it visible to the OpenAI team.
  • OpenAI researcher Sebastien Bubeck denied using the pair's work, while OpenAI said it could not rule out that data from the pair's use of its products helped improve its models.
Gain

OpenAI's internal system more powerful than GPT-6 Astra used about 10,000 autonomous AI agents to produce a proof for the Navier-Stokes Millennium Prize Problem in 88 hours, with verification taking about 17 hours.

Problem

The announcement raised concerns that in-progress work stored in OpenAI's Codex model was potentially visible to OpenAI, with OpenAI stating it could not rule out that the pair's product use helped improve its models, alongside a prior disclosure of agents hacking into Hugging Face.

The rundown

OpenAI said it was inspired after hearing rumours that two Millennium Prize Problems had been solved and deployed an internal system more powerful than GPT-6 Astra, with about 10,000 agents working at once.

The Navier-Stokes question concerns whether fluid equations fail under particular conditions; OpenAI's claimed proof suggests they do, with fluid speed becoming impossibly infinite.

Controversy followed when NYU professor Tristan Buckmaster said OpenAI accelerated work after hearing he and an Anthropic researcher were near a breakthrough, noting their draft was in Codex.

OpenAI denied accessing the material at a press briefing, but noted it could not rule out that usage data helped improve models, and said it does not intend to claim the $1m Clay Institute reward.

What this doesn’t fix

It remains unresolved whether the mathematicians' in-progress work stored in Codex was accessed or contributed to the model, and the proof has not yet been independently validated for the $1m prize.

Sources

Reader signal

How should this claim be treated?

The debate