OpenAI announced that its AI solved a 100-year-old mathematical challenge, but controversy is mounting as mathematicians allege the AI was trained on and plagiarized their unpublished research results.
Imagine this: You have spent years agonizing over a problem, working day and night to solve a challenge that no one in the world has ever cracked. Just as you are about to announce your findings to the world, a massive IT corporation suddenly declares, “We solved the problem using AI.” How would you feel?
The mathematical community is currently pushing back fiercely over this very situation. Alongside the expectation that artificial intelligence will become a tool to accelerate human intellectual achievement, strong questions are being raised about whether the process of creating such tools is transparent.
Why Does This Matter?
This controversy is not just about crowning a winner in a “math problem solving” contest. It is a matter of “research ethics” and “data sovereignty”—specifically, how AI learns from and utilizes the data we produce. If AI is secretly harvesting the unpublished knowledge of researchers, who would ever want to share or publish their valuable findings freely again? This is a serious issue that could ultimately stifle the intellectual curiosity of all humanity. [Source: OpenAI claims to have solved math’s greatest unsolved problems, but mathematicians allege it played dirty (Live Science)]
Making It Simple: Culinary Secrets and Recipes
You can compare this complex situation to a “culinary secret.”
Suppose a chef has spent decades researching ingredient combinations to create their own secret sauce recipe. Just before the chef is about to announce the recipe to the world, a giant food factory advertises that its state-of-the-art automatic cooking machine has produced a dish with the exact same perfect flavor. Naturally, the chef would suspect that the machine stole their secret recipe.
In this incident, mathematicians suspect that OpenAI used their unpublished materials—which they were still in the process of researching—as training data for its AI. [Source: Mathematicians want proof OpenAI didn’t use their work (The Verge)] [Source: The drama over OpenAI’s math breakthrough claim (Business Insider)]
Of course, there are foundational principles commonly used in solving mathematical problems. In the case of this Navier-Stokes equation challenge (equations that explain fluid flow in fluid dynamics), both the mathematicians and OpenAI reportedly used an approach pioneered by Diego Córdoba and Luis Martínez-Zoroa. [Source: What OpenAI’s latest controversy tells us about the future of math (Technology Review)] However, the mathematicians remain unconvinced, arguing that the decisive process used to reach the result is too similar to their own research to be a coincidence.
Where We Stand Now
On September 8, 2026, OpenAI announced that it had used its model to solve a mathematical challenge that has perplexed humanity for the last 100 years. [Source: OpenAI claims to have solved math problem that stumped humans for decades (The Guardian)] The problem was significant enough to carry a $1 million prize. [Source: Credit dispute sparked by math breakthrough claim (USA Today)]
In response, OpenAI’s Sebastien Bubeck forcefully refuted the allegations, stating, “We did not use their prompts or proofs in our model,” and characterized the claims as “false and inflammatory.” [Source: OpenAI claims to have solved math’s greatest unsolved problems, but mathematicians allege it played dirty (Live Science)] Nevertheless, criticism continues to mount among mathematicians regarding the lack of transparency in OpenAI’s data training process. [Source: Another mathematician accuses OpenAI of unethical behavior (Jingletree)]
What’s Next
The trust between the mathematical community and the AI industry has been deeply damaged. The important takeaway is that demands for transparent disclosure of the “process” by which AI models derive their answers will only grow louder. Furthermore, researchers will likely become much more proactive in protecting their intellectual property rights—either by sharing data selectively or by actively exercising their right to refuse AI training.
As AI advances, we are entering an era where the “legitimacy of the process” is just as important as the “result.” What we need to pay attention to is not just which challenge the AI has solved, but whether the path to that solution is a result that respects humanity’s collective effort.
AI Opinion
The issue of data transparency is an unavoidable topic in the age of AI. Establishing fair compensation and ethical standards during the process is just as important as the technological progress itself. This incident reminds us once again that what is more important than the speed of technological advancement is the trust that sustains it.
References
-
[Mathematicians want proof OpenAI didn’t use their work The Verge](https://www.theverge.com/ai-artificial-intelligence/993263/where-does-openai-get-mathematics-training-data) - Mathematicians want proof OpenAI didn’t use their work - Wilson’s Media
- Mathematicians want proof OpenAI didn’t use their work - Jingletree
-
[Mathematicians want proof OpenAI didn’t use their work Hacker News](https://news.ycombinator.com/item?id=49641792) -
[‘Why would you ruin your career?’: OpenAI claims to have cracked one of math’s greatest unsolved problems, but mathematicians allege it played dirty Live Science](https://www.livescience.com/physics-mathematics/mathematics/why-would-you-ruin-your-career-openai-claims-to-have-cracked-one-of-maths-greatest-unsolved-problems-but-mathematicians-allege-it-played-dirty) - OpenAI’s Big Math Breakthrough Claim Sparks Drama - Business Insider
- Mathematicians want proof OpenAI didn’t use their work
- OpenAI Says It Solved Math’s Deepest Problem But …
- OpenAI claims blockbuster math breakthrough amid swirl of …
- What OpenAI’s latest controversy tells us about the future of …
- OpenAI claims to have solved maths problem that stumped …
- Math breakthrough claim by OpenAI sparks credit dispute
- Can OpenAI’s math breakthrough be trusted? Mathematicians cry foul
- OpenAI researcher allegedly pressured mathematician to drop…
- Did OpenAI race ahead on Navier-Stokes? - News9live
- The AI made a calculation error
- The AI used their unpublished research without authorization as training data
- The research costs were too high
- They admitted the facts and apologized
- They immediately scrapped the research
- They refuted the claims as baseless, stating their model did not use unpublished data
- The Navier-Stokes equations problem
- Fermat's Last Theorem
- The Riemann Hypothesis