AI coding assistants have changed the inner loop — code is being written faster than ever. But most delivery pipelines, reliability practices, and governance frameworks were built for human-authored code. This session examines what actually breaks, what gets better, and what entirely new problems emerge when the primary author is an AI. From rethinking CI/CD validation to governing AI-generated contributions in regulated industries, these leaders share what software delivery actually looks like when the machine does the heavy lifting.

Internal developer platforms were built for human developers. Now AI agents are consumers of those same platforms — with completely different needs, failure modes, and trust requirements. This session explores what platform engineering looks like when you're building for both: how the platform team's mandate is shifting, what 'good developer experience' means when AI handles more of the build, and how leading teams are redesigning their internal platforms to serve humans and agents without compromising either.

‍

Every enterprise says they're doing AI. Far fewer have answered: how do you govern it, build genuine fluency across the organization, and make deliberate decisions about speed versus safety — especially when regulators, boards, and customers are watching? This session brings together leaders from regulated industries solving this differently: one restructuring the business around AI from the top down, one taking a deliberate governance-first approach, and one navigating the tension between AI speed and airline-grade reliability standards.

AI-generated code is outpacing security teams' ability to review it. The attack surface is expanding faster than manual processes can track, and the old model — security as a gate at the end of the pipeline — is no longer viable. This session examines what it takes to embed security continuously: from shifting left on vulnerability detection to building the organizational muscle for real-time governance on AI-generated contributions. What does DevSecOps look like when your developers build with AI copilots and your adversaries hunt with AI agents?