When AI can identify hundreds of system vulnerabilities in just hours, governance operating at human speed is dead in the water. In this episode of Agents of Dev, Mitch Ashley and Brad Shimmin break down critical insights from recent AWS and Red Hat enterprise briefings, exploring why conventional pipelines fail under agentic workloads. They dive into AWS's findings on right-sizing guardrails to the actual cost of failure, Red Hat’s Lightwell project for autonomous vulnerability patching across open-source packages, and the security liabilities introduced by unmanaged agent configuration files. Finally, they look ahead to KubeCon to preview how containerization and sandbox runtimes must evolve to host the next generation of autonomous agents. This and more on Agents of Dev, part of The Futurum Media Podcast Network.