AI Safety, geometric analysis, metric geometry, low-dimensional topology.
AI safety for mathematicians
Click here to sign up
for the mailing list on AI safety events at the Fields Institute and the U of T math
department. The link opens a pre-addressed email; send it (it can be
blank), then reply to the confirmation message the LISTSERV sends back.
I am accepting Master's and PhD students who want to work on mathematical
aspects of AI safety, control and alignment.
In the fall of 2026 I am teaching a graduate course on
Mathematics for AI Safety
at the Fields Institute.
Mathematical AI Safety InstituteMathematical AI Safety Institute is seeking mathematicians to relocate
to the Bay Area and begin research relevant for reducing catastrophic and existential
risk from AI. Apply here: https://maisi.org/apply
Fields–PrincInt postdoctoral fellowships in Mathematics for AI SafetyTwo-year Fields Postdoctoral Fellowships, run by the Fields Institute
in partnership with Principles of Intelligence, supporting research at the
intersection of mathematics and AI safety. Research directions include
rigorous mathematical
foundations of deep learning for AI interpretability, formal verification
methods, modeling emergent risks from AI R&D automation. Part of the
Fields Institute Centre for Mathematical AI;
apply via MathJobs.
AI Safety for MathematiciansAn introduction to the field written for working mathematicians,
maintained by Jacob Tsimerman.
ILIAD IntensiveA four-week, full-time course on foundational AI alignment research
for mathematicians, theoretical physicists and theoretical computer scientists.
Tuition is free and comes with a travel and housing allowance; run in London and
Berkeley.
Toronto AI Safety InitiativeRuns a fellowship and an intensive course training for students in
Toronto to become AI safety researchers. A great place for undergraduate students
to begin their research in AI safety.
Trajectory LabsA Toronto AI safety office providing workspace, talks and a peer
network for people working in the field.