AI Resilience Report forMathematicians

Right now, AI is mostly augmenting mathematicians rather than replacing them — but the change is happening fast. Mathematicians who once dismissed AI models as too error-prone started playing around with them, and those early adopters found that the models could help break genuinely new ground, using AI to discover and prove new results, accomplishing in a day what would have once taken weeks or months. Fields Medalist Terence Tao explains the split this way: "I don't think AI will replace mathematicians, but it will complement them.

There could be a division of labour: we decide what to prove and what we think is interesting. We could get instant feedback from the AI." Concrete examples are piling up. A 2026 collaboration between humans and AI formally verified Maryna Viazovska's Fields Medal–winning sphere packing proofs, signaling rapid progress in AI's abilities to assist with mathematical research.

A February 2026 "First Proof" contest gave AI models grad-school-level questions ^[1], and with varying levels of autonomy, the models succeeded in solving over half the problems — the kinds of tasks that map directly onto the "perform computations" and "develop models" parts of a mathematician's job.

Adoption in math is moving unusually quickly, and the SIAM AI Task Force Report published in March 2026 ^[2] argues applied mathematics is essential infrastructure for the future of AI. One reason is technical: in mathematics, almost uniquely, you can automatically check the output, so AI companies have recognized that their most unambiguous successes are going to come from mathematics. Tools are also cheap and everywhere — ChatGPT, Claude, and Gemini all work on math, and some mathematicians are now leaving academia to work at big tech firms like OpenAI and Google, or to join math-focused AI startups such as Harmonic, Logical Intelligence, Axiom Math, and Math Inc.

The job market is still strong: the U.S. Bureau of Labor Statistics projects ^[3] that the growing adoption of AI technologies, including generative AI tools, will fuel strong job growth among computer and mathematical occupations, with mathematical science roles expanding fast. But adoption isn't frictionless — a March 2026 Harvard Business Review analysis ^[4] tracks how AI is reshaping knowledge work, and within math itself attitudes are very much a spectrum, with all the "five stages of grief" — denial, anger, bargaining, depression and acceptance — playing out. The honest takeaway for students: human judgment about what's worth proving, creativity, and the ability to communicate ideas still matter a lot.

Those who understand maths traditionally but are also adept at using new tools can flourish — so learning both is the smart bet.