Fields Medalist who warned AI could threaten humanity joins OpenAI

"I'll be starting a position at OpenAI," Tsimerman, a professor at the University of Toronto, said at a news conference following the Fields Medal award ceremony in July, as quoted by The Wall Street Journal.

Rather than leaving mathematics behind, Tsimerman said he sees the move as applying mathematical rigor to what he considers one of the world's most important scientific challenges.

"In a few years, AI systems will be robustly superhuman at the act of doing mathematics," he said. "The social consequence of that, how we choose to react, what you feel about it—those are much harder questions."

He plans to use mathematical verification methods to better understand how advanced AI systems make decisions and help ensure they behave safely as their capabilities continue to improve.

Tsimerman told the Toronto Star that advances in AI coding capabilities had significantly reduced the disadvantage of not having a software engineering background, allowing him to begin running experiments and contributing to AI safety research much more quickly than even a year ago.

He also said joining OpenAI would allow him to build expertise in machine learning while working alongside researchers tackling the same challenges.

"Because the biggest hole in my background is the machine learning, I thought going to a lab made the most sense at this moment," he said.

Although leading AI companies can offer compensation that universities cannot match, Tsimerman said money was not the deciding factor. "My main thing was where I could position myself better," he said.

2026 Fields Medal winner Jacob Tsimerman. Photo from University of Toronto website

Tsimerman, along with three other mathematicians, received the Fields Medal, widely regarded as the Nobel Prize of mathematics, at the International Congress of Mathematicians in Philadelphia on July 23.

He became the first scholar based at a Canadian institution and only the second Canadian to receive the award since it was established in 1932. He was recognized in part for helping prove the André-Oort conjecture, a longstanding problem in arithmetic geometry.

His decision to join OpenAI surprised many in the mathematics community because he has been among the field's most outspoken voices warning about the risks of advanced AI.

For much of his career, Tsimerman remained skeptical that AI could replicate uniquely human abilities such as intuition and creativity. However, after experimenting with ChatGPT in 2022, he said his outlook changed as he became convinced the technology would advance far more quickly than he had expected.

Last year, Tsimerman co-authored a paper examining scenarios in which AI could lead to humanity's extinction. "Killed off as pests," he writes, "by intellectually and physically superior machines."

He also stopped taking graduate students who were not engaging with AI because he was uncertain about the future of mathematics, and shifted his own research away from number theory and complex algebraic geometry.

His concerns have grown alongside rapid advances in AI's mathematical abilities. OpenAI recently said one of its unreleased models solved an 80-year-old mathematical problem, while Anthropic's Fable model identified a counterexample to the Jacobian conjecture, a problem that had challenged mathematicians for 87 years. The researcher who reported the discovery was one of Tsimerman's former doctoral students.

With his last Ph.D. students graduating, Tsimerman said it was a natural time to begin a new chapter. He does not know how long he will remain at OpenAI or when he will return to teaching.

"There are very few certainties or permanent decisions in my life right now," he said.

Despite the career change, Tsimerman said mathematics remains central to his future. "I have no shortage of things," he said. "Math is one of them."