Sessions / Menlo Security: The AI Agent Security Misdirection: Why the Industry Is Fighting the Wrong Battle
Menlo Security: The AI Agent Security Misdirection: Why the Industry Is Fighting the Wrong Battle
Wednesday, June 03, 2026 / 11:15 AM - 11:45 AM EDT
Share
About the Session
AI agent security has converged on a familiar set of battles — model hardening, AI gateways, identity governance, red-team evaluations, shadow AI. But most of them miss where the real risk now lives: in runtime legitimate use, sanctioned permissions, and machine-speed actions no current control was designed to see. A candid look at what the industry is getting wrong about agent risk and what an honest agenda for the next few years should include.