AI has moved fast and so has the magnitude of your worst day. In just 12 to 24 months, enterprise AI has evolved from individual experimentation into the early era of autonomous, agentic systems running across endpoints, cloud, and data center. That shift has expanded the security conversation from a narrow guardrails problem into something far broader: discovering where AI lives, understanding agent intent and actions, and unifying it all under a common control plane.

This session traces that journey, covering where we have been, what we have today, and where we are headed. We will start by framing how the AI risk landscape has changed and why traditional tools fail against non-deterministic, machine-speed systems. From there, we will walk through how AI security platforms are evolving with AI discovery as the essential starting point and how AI security testing and runtime controls must be intertwined. We will close with a look ahead at how AI Gateways are becoming the enterprise control layer, and how governing model routing, token usage, and agent behavior across the organization will become future table stakes for AI risk leaders. (Presented by F5)