Agentic AI is moving from “chat” to “do”—systems that can plan, take actions across tools, and iterate toward an outcome with less step-by-step human prompting. Join Irene Liu, Executive Director of Stanford Law School’s AI Initiative, for a conversation with Morrison Foerster’s Justin Haan on where agentic AI stands today, what it can (and can’t) reliably do, and how it may reshape the practice of law.