AI Security: First Principles

Columbia CS professor and AI security researcher. AI safety beyond panic and compliance checklists: building technical defenses that actually make systems safer.