AI Doesn't Break Security. It Exposes It. | Jason Garbis

Most security breakthroughs don't come from new technology, they come from finally getting the basics right. As AI accelerates both attacks and the pace of vulnerability disclosure, the organizations that fare best aren't the ones chasing the next big thing, but the ones with the visibility, governance, and segmentation to absorb the shock. In this episode, Raghu Nandakumara sits down with Jason Garbis, founder and CEO of Number Line Security and co-chair of the Zero Trust Working Group at the Cloud Security Alliance, to explore what Zero Trust looks like when threats evolve at machine speed. Jason draws on his path from early software-defined perimeter work to leading Zero Trust strategy at the Cloud Security Alliance, sharing why "right-sizing" a Zero Trust initiative matters more than chasing sweeping transformation, and why even a well-built security capability fails without genuine buy-in from the business. The conversation then turns to AI's effect on the threat landscape: the surge of vulnerabilities and patches enterprises now have to absorb, the widening gap between attacker speed and defender response, and why segmentation remains one of the most effective ways to shrink the blast radius of an attack. Raghu and Jason discuss: How to right-size a Zero Trust initiative for an organization's actual readiness Why "build it and they will come" doesn't work for security adoption Tying Zero Trust investments to business priorities like AI adoption, M&A, and compliance What's genuinely new — and what isn't — about securing AI systems and agents How accelerating vulnerability disclosures are reshaping patch management Why segmentation reduces blast radius even when patching can't keep pace A simple analogy for explaining zero trust to non-security stakeholders Jason closes with his go-to way of explaining Zero Trust to non-technical stakeholders: an analogy involving brakes, oil, and seat belts that reframes security as a shared responsibility rather than a roadblock. See more episodes of The Segment here: https://social.illumio.com/ug1ch Explore more resources from Illumio: Learn about Illumio: https://www.illumio.com Learn about Breach Containment: https://www.illumio.com/breach-contai... Read the latest cybersecurity insights: https://www.illumio.com/blog Explore The Zero Trust Hub: https://hub.illumio.com/ Follow Illumio on Facebook:   / illumio   Follow Illumio on Instagram:   / illumioinc   Follow Illumio on LinkedIn:   / illumio   Follow Illumio on X: https://x.com/illumio Episode Chapters 01:13 — Jason's Path into Zero Trust 03:26 — From SDP to Zero Trust: Key Industry Shifts 05:07 — Is Zero Trust Adoption Held Back by a Lack of Discipline? 08:08 — How Zero Trust Adoption Has Evolved Over Five-Plus Years 10:46 — Right-Sizing Zero Trust: When Smaller Is the Right Answer 13:18 — What Good Zero Trust Programs Get Right 16:46 — Where Zero Trust Programs Go Wrong 20:26 — Why "Build It and They Will Come" Doesn't Work 23:01 — Securing AI Systems: Building a Taxonomy 24:34 — Is AI a New Problem, or an Old One at Higher Speed? 26:38 — Reacting to Mythos: The Coming Patch Tsunami 31:26 — The Attacker's Real Advantage: Speed of Exploitation, Not Discovery 32:48 — The Widening Asymmetry Between Attackers and Defenders 33:45 — The Cloud Security Alliance's Mythos-Ready Security Guidance 37:45 — What Zero Trust Strategy Looks Like Post-Mythos 40:10 — Making Security a Byproduct of Operations 42:33 — Jason's Zero Trust Analogy: Brakes, Oil, and Seat Belts #ZeroTrust #AISecurity #CyberResilience #CyberSecurity #SecurityLeadership #Segmentation #VulnerabilityManagement #CloudSecurity