In this episode, James Maude sits down with Jeffrey Matson, CEO of SecureAuth, whose thirty year career runs from encrypting early GPS satellites for the military to building the ASICs behind deep packet inspection and leading identity security at Huawei, Juniper Networks, MistNet, and Zage Security. Jeffrey explains why AI agents create a brand new category of risk, capable of doing catastrophic damage without ever being hacked, and shares the story of a CISO whose own vibe coding executive became the company's biggest insider threat. They dig into the quiet authorization failures inside MCP servers, why identity has gone from cybersecurity's least glamorous discipline to the center of the action, and why the industry's obsession with the perimeter is finally giving way to real zero trust. Jeffrey also draws a sharp parallel between the dot com bubble and today's AI investment boom, and offers advice for anyone trying to break into cybersecurity in the age of agentic AI.