Jacob Tsimerman will trade the blackboard for a model lab. The University of Toronto professor, one of four mathematicians awarded the 2026 Fields Medal at July’s congress in Philadelphia, said at a news conference that he is joining OpenAI to work on AI safety.
Tsimerman does not see the job as leaving mathematics behind. He argues it is about applying mathematical rigor to what he regards as one of the era’s most important scientific problems, and he plans to lean on formal verification to understand how advanced AI systems make decisions and keep them safe.
The hire stands out because Tsimerman has been one of the field’s loudest voices warning about advanced AI, including a 2025 paper sketching “omnicide events” in which AI could contribute to human extinction. He has also said AI will soon outperform humans at mathematical research, and he credits AI coding tools with removing the software-engineering barrier that once kept him from running experiments.
For most of his career Tsimerman doubted AI could replicate human intuition and creativity, making the appointment all the more striking. His exit adds to the flow of elite researchers into the biggest labs, where safety teams compete for exactly this kind of talent.
