What We Learned Securing AI Agents at Scale on Multi-Tenant Kubernetes Clusters
Talkat KubeCon + CloudNativeCon + OpenInfra Summit + PyTorch Conference China 2026 · 2026-09-07..09-09·
summary
Describes a defense-in-depth architecture for running hundreds of personal AI agents on a shared Kubernetes cluster using Kata Containers for VM isolation, Agent ID via OIDC federation to give each agent a unique cloud identity with scoped access, NetworkPolicy to block IMDS and lateral movement, ValidatingAdmissionPolicy to stop exec breakouts, and governance guardrails with audit logging to enforce blast-radius limits.
AI-written summary · about
related talks
42mZero Trust for Autonomous Agents: Isolating AI Workloads on Kubernetes
25mYour AI Is a Toddler With Root Access: Why Kubernetes Policies Are Your Last Line...
30mSo You Want To Run AI Agents on Kubernetes: A 101 Guide
21mSecurity Best Practices for AI on Kubernetes
23mAgents in Kubernetes 101: A Crash Course on Cloud Native AI Agents
25mSecuring AI Pipelines: Real-World Attacks on Kubernetes-Based AI Infrastructure
40mIn AI We Trust? Securing the Future, One Agent at a Time
11mLightning Talk: Your Most Privileged User Isn't Human: The Agent Security Blinds...