# Mathematicians hate AI. They can't quit it.

> Top mathematicians accuse AI companies of stealing their work, but they keep using the tools anyway.
- Title: Mathematicians hate AI. They can't quit it. · WIRED | Shortify
- Summary: Top mathematicians accuse AI companies of stealing their work, but they keep using the tools anyway. Mathematicians feel trapped between distrust of AI…
- Keywords: mathematicians, openai, anthropic, ai ethics, attribution, navier-stokes problem, tristan buckmaster, andreas thom, ai training data, academic integrity, mathematical research, hate
- Source: WIRED — https://wired.com/story/mathematicians-cant-quit-ai
- Author: Isabella Ward
- Published: 2026-09-19T10:00:00+00:00
- Read time: 2 min
- Topics: mathematicians, openai, anthropic, ai ethics, attribution, navier-stokes problem, tristan buckmaster, andreas thom, ai training data, academic integrity, mathematical research
## Mathematician accuses OpenAI of theft
NYU professor Tristan Buckmaster claims OpenAI used his work to beat him to solving the Navier-Stokes problem, a legendary math challenge with a $1 million bounty.
## But he still uses their tools
Despite his accusations, Buckmaster continues using OpenAI's Codex to edit papers and understand the company's mathematical reasoning. He says he's trapped by AI's usefulness.

> These companies have a monopoly, and there is not much choice.
## OpenAI denies copying his work
The company investigated and amended its announcement, stating that Buckmaster's prompts from the two months before the September 8, 2026 paper could not have influenced the system.
## Another mathematician raises concerns
German mathematician Andreas Thom, who spent two decades developing techniques in geometric group theory, was shocked when OpenAI claimed its model used them without crediting his work.
## OpenAI amended that announcement too
After Thom pointed out that OpenAI overlooked his 2019 paper and other mathematicians' contributions, the company updated its press release to acknowledge prior work.
## Attribution becomes impossible
Thom says AI makes it impossible to trace who contributed what to breakthroughs, upending how science has traditionally worked with peer review and proper credit.

> AI really kills this entire idea that you could trace back who contributed what.
## Mathematicians face an existential question
Cornell mathematician Alex Townsend asks how individual researchers can contribute when trillion-dollar companies with massive computing power are competing in the field.
## The pressure to adopt AI is real
Townsend reports that colleagues are now asking how to access more powerful AI models, driven by both excitement about capabilities and anxiety about being left behind.

> I feel both excited and nervous simultaneously.
## Even skeptics keep using the tools
Thom continues using ChatGPT with privacy settings to speed up paper writing, accepting that his field is changing even as he worries about attribution.
## Some mathematicians are pushing back
Twenty-five Fields medalists signed an open letter about misalignment between AI companies and mathematicians. Over four thousand people signed the Leiden Declaration calling for safeguards.
## Buckmaster calls for ground rules
He's proposing a temporary pause on AI-driven math breakthroughs while the community and AI labs establish guidelines for proper attribution and result release.
## The next generation needs guidance
Students are asking what their future looks like as AI becomes capable at solving math problems. Mathematicians worry early-career researchers will be isolated if they don't adopt the technology.
## Key takeaway

Mathematicians feel trapped between distrust of AI companies and the competitive pressure to use their tools or fall behind.