OpenAI claims to have solved the 90-year-old Navier-Stokes problem using 10,000 AI agents, but it is facing friction with the math community over allegations of data misappropriation.
Imagine a puzzle that has haunted the world’s most brilliant mathematicians for the past 90 years. This problem is so difficult that anyone who solves it is eligible for a massive $1 million prize. Then, one day, an AI company announces, “Our AI army solved it in 88 hours.” Would people be thrilled, or bewildered?
The recent news that OpenAI solved the ‘Navier-Stokes’ equation has ignited intense debate in the mathematics community rather than celebration. Has AI truly transcended human mathematical limits, or has it just swiped someone else’s hard work?
Why does this matter?
| The ‘Navier-Stokes equation’ is a fundamental formula that describes the movement of fluids (water, air, etc.). It is essential for understanding how airplanes fly or how typhoons move. The reason it was designated as one of the ‘Millennium Prize Problems’ is that proving whether the solutions to these equations always exist and are smoothly defined is one of the greatest homework assignments in modern mathematics. [Source: OpenAI fought dirty on career-making math problem… | TechCrunch](https://techcrunch.com/2026/09/08/openai-fought-dirty-on-career-making-math-problem-says-nyu-mathematician/) |
If AI truly solved this, it would be a monumental achievement in human scientific history. However, as allegations of ‘misappropriating’ other mathematicians’ research results have surfaced, the ethical issues of AI technology and the transparency of data provenance have once again come under scrutiny.
Understanding it easily
In simple terms, OpenAI essentially hired 10,000 ‘math assistants’ to run calculations non-stop. Here, the 10,000 assistants refer to AI agents capable of performing tasks on their own. Source: Controversy erupts as OpenAI claims solution to Navier-Stokes maths…
| To use an analogy, it’s like 10,000 people splitting the work of building a complex Lego castle and fitting the blocks together for 88 hours without a break. [Source: OpenAI Says 10,000 AI Agents Solved Navier-Stokes… | Yellow](https://yellow.com/news/openai-10000-agents-navier-stokes-dispute) To verify whether the results of this massive undertaking were mathematically correct, they used a program called ‘Lean’. Lean is a language that allows a machine to self-verify whether the results produced by an AI are logically airtight. Source: OpenAI’s AI solved a math prize problem. The mathematician says it… |
But what if these ‘assistants’ had sneaked a peek at a researcher’s notebook and copied the answer? Mathematician Tristan Buckmaster strongly suspects that the AI may have secretly referenced his research notes during its training process. Source: A Mathematician Working On The Navier-Stokes Millennium Prize…
Where do we stand?
OpenAI has officially denied the theft allegations, stating, “Our research team and agents had no access to external research results before they were made public.” Source: A Mathematician Working On The Navier-Stokes Millennium Prize…
| However, the controversy is not easily dying down. While it was announced that the AI produced the results in 88 hours and verification took an additional 17 hours, exactly what data was used and with what logic remains shrouded in mystery. [Source: OpenAI Says 10,000 AI Agents Solved Navier-Stokes… | Yellow](https://yellow.com/news/openai-10000-agents-navier-stokes-dispute) Mathematicians are still demanding a clear explanation regarding data provenance, leading to a tense standoff with OpenAI. Source: The mathematicians published machine-checkable proofs. OpenAI… |
What lies ahead?
This incident has opened a ‘new era’ where AI solves mathematical challenges, while simultaneously posing fundamental questions about what ‘academic integrity’ means. In the future, it is highly likely that transparency regarding the ‘provenance of data’ used in the process—not just verifying whether the AI’s answer is correct—will become a new standard in the scientific community.
We now live in an era where we must go beyond asking whether AI is smart and instead ask, “Whose knowledge did it borrow to become smart?” Only when technological progress can prove the legitimacy of its process as much as its achievements can it be called true scientific advancement.
A word from the AI
AI’s mathematical achievements are undoubtedly impressive. However, the core of mathematics is not just finding the answer, but the perfect logical process that leads to it. Unless the suspicions regarding data provenance are clearly resolved, this achievement will inevitably remain a subject of verification rather than praise.
References
- Controversy erupts as OpenAI claims solution to Navier-Stokes maths…
-
[OpenAI fought dirty on career-making math problem… TechCrunch](https://techcrunch.com/2026/09/08/openai-fought-dirty-on-career-making-math-problem-says-nyu-mathematician/) - A Mathematician Working On The Navier-Stokes Millennium Prize…
-
[OpenAI Says 10,000 AI Agents Solved Navier-Stokes… Yellow](https://yellow.com/news/openai-10000-agents-navier-stokes-dispute) - OpenAI’s AI solved a math prize problem. The mathematician says it…
- The mathematicians published machine-checkable proofs. OpenAI…
- OpenAI claims to have solved Navier-Stokes math problem
- Quantum computer
- 10,000 AI agents
- Traditional supercomputer
- Fields Medal Problems
- Millennium Prize Problems
- Turing Problems
- AI calculation errors
- Misappropriation of personal research data
- Prize money distribution issues