Editorial illustration for OpenAI Investigates Mathematician's AI Obsolescence Claims
Mathematician Claims AI Solved Million-Dollar Problem
OpenAI Investigates Mathematician's AI Obsolescence Claims
Tristan Buckmaster spent months chasing a solution to the Navier-Stokes existence and smoothness problem, one of math's seven Millennium Prize puzzles carrying a $1 million bounty. He used OpenAI's Codex and Anthropic's Claude to help with the work, alongside Anthropic researcher Levent Alpöge. Then OpenAI announced it had solved the same problem, deploying tens of thousands of agents to get there. Buckmaster says the timing wasn't a coincidence: he believes OpenAI got wind of how close he was and used that knowledge to sprint past him.
The NYU professor went public with the accusation about a week and a half ago, and it set off an argument far bigger than one disputed proof. Suddenly people were asking whether AI is about to make working mathematicians unnecessary, and whether the field's biggest companies are willing to cut corners to prove it. OpenAI has since opened its own investigation into what happened.
What makes the situation stranger is Buckmaster's own behavior since the accusation. He hasn't stopped using OpenAI's tools. He's been running Codex to clean up his papers and even to reconstruct how OpenAI's agents might have followed his earlier work toward their final answer.
“Even if you don't agree with any of this, you're kind of stuck. With AI being so useful, it's hard to completely prevent oneself from using it,” Buckmaster tells WIRED.
Why this matters
OpenAI's decision to amend its Navier-Stokes announcement after Buckmaster's complaint is a small but telling admission: the company acknowledged his Codex work played a role it hadn't originally credited. That matters more than the math problem itself. If a lab can quietly reshape a public claim of scientific credit only after a researcher goes public, what happens to the mathematicians who don't have WIRED on speed dial?
For developers and founders building on these models, Buckmaster's situation is a preview of a coming dispute over attribution, not just for prose or code, but for mathematical proofs and research priority. His own admission, that he can't stop using the tools even as he distrusts the company behind them, describes the market position OpenAI and its rivals now hold. That's the real story here: a handful of firms control the infrastructure researchers depend on, and dependency is doing a lot of the work that trust used to do.
Anyone building products on top of frontier models should watch how OpenAI handles credit disputes, because it's a preview of how it'll handle yours.
Related Reading
Browse all OpenAI news →
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
Related Topic OpenAI researcher quits, citing distrust over ad‑driven engagement metrics
LLMs & Generative AI
Related Topic OpenAI launches GPT-Image 1.5 with precise editing for enterprise visuals
LLMs & Generative AI
From Archives Manus Targets USD 4B Valuation in USD 500M Fundraise After Meta Separation
From Archives Salesforce Agentforce Generates Synthetic Tests to Pressure-Test AI Agents
From Archives Researchers Use Corrupted Image to Hack OpenAI
Open Source
From Archives Vercel Engineer Uses OpenAI's ChatGPT Luna 5.6 to Review Commands
LLMs & Generative AI
Common Questions Answered
What is the Navier-Stokes existence and smoothness problem that Tristan Buckmaster was working on?
The Navier-Stokes existence and smoothness problem is one of mathematics' seven Millennium Prize puzzles that carries a $1 million bounty for its solution. Buckmaster spent months attempting to solve this problem using OpenAI's Codex and Anthropic's Claude alongside researcher Levent Alpöge before OpenAI announced it had solved the same problem using tens of thousands of AI agents.
Why does Buckmaster believe OpenAI's timing in solving the Navier-Stokes problem was not coincidental?
Buckmaster suspects that OpenAI became aware of how close he was to solving the Navier-Stokes problem and used that information to prioritize their own solution effort. His concern centers on the possibility that OpenAI leveraged knowledge of his progress to deploy their own resources strategically, raising questions about research ethics and credit attribution in AI-assisted mathematics.
What did OpenAI acknowledge after Buckmaster's public complaint about the Navier-Stokes announcement?
OpenAI amended its original Navier-Stokes announcement and acknowledged that Buckmaster's Codex work played a role in the solution that it had not originally credited. This amendment represents an admission by the company that proper scientific credit attribution was missing from their initial public claim.
What is the broader concern Buckmaster raises about mathematicians and AI tool usage?
Buckmaster highlights a dilemma where mathematicians feel compelled to use AI tools like Codex and Claude because they are so useful and effective, yet doing so creates risks around credit attribution and potential exploitation of their work. He notes that even those skeptical of AI struggle to completely avoid using these tools given their practical advantages in research.
Why does the article suggest OpenAI's amendment to its Navier-Stokes claim matters beyond the math problem itself?
The amendment raises concerns about how AI companies handle scientific credit attribution and whether researchers without public platforms can protect their contributions. The incident demonstrates that if a lab can quietly reshape public claims of scientific credit only after pressure, there may be systemic issues affecting mathematicians who lack media visibility to challenge such decisions.
Further Reading
- OpenAI claims to have solved maths problem that stumped humans for decades - The Guardian
- Math breakthrough claim by OpenAI sparks credit dispute - USA Today
- OpenAI says it cracked 90-year-old maths problem in 88 hours - BBC News
- AI Has Solved One of Math's $1 Million Millennium Prize Problems - Quanta Magazine
- OpenAI claims huge maths breakthrough on a famed Millennium Problem - Nature
