Securing the Invisible Hands - Policy and Guardrails for AI-Driven Infrastructure

Explore how to secure AI-driven infrastructure automation through policy enforcement and guardrails in this 42-minute conference talk from the Linux Foundation. Discover architectural patterns for maintaining security control while enabling AI agents to automate infrastructure decisions from provisioning to scaling without human oversight. Learn how Crossplane 2.0's declarative control plane enables secure AI-driven automation through policy enforcement, real-time monitoring, and isolated execution environments. Examine practical approaches for enforcing compliance through composition functions, isolating AI workloads with network policies, and auditing machine-generated infrastructure changes. See how Crossplane 2.0's enhanced reconciliation engine detects security drift within seconds and automatically implements remediation through live demonstrations. Understand how composition functions respond to security events, enforce just-in-time access, and implement dynamic compliance policies. Gain insights into real-time workload isolation patterns essential for organizations deploying AI-managed infrastructure. Walk away with open source strategies for maintaining security control while enabling AI automation at cloud scale using vendor-neutral tools and technologies.