University of Toronto institute whose work includes interpretability, alignment and oversight research alongside technology governance. Named initiatives include "An End-to-End Approach to Safe and Secure AI Systems" (a $5.6M NSERC-funded project led by David Lie), David Duvenaud's Post-AGI workshop on societal change after transformative AI, and an AI Safety discussion group led by Sheila McIlraith and Toryn Klassen.