AI holds great promise and, many believe, great peril. What can mathematicians contribute to ensuring that promise is fulfilled, and peril avoided? I’ll highlight some research directions to be explored in my topics course this fall, MATH 7710: Math for AI Safety. I’ll also touch on the question of how to empower students to succeed in a world pervaded by AI, and to shape that world for the better: After all, many of today’s math majors will be tomorrow’s AI engineers. This talk will be interactive, and I’ll mostly supply questions rather than answers, so please come prepared to debate!