Securing AI Agents: From Code to Production

This DevOps Vancouver Meetup covers the security and governance of AI agents throughout the software lifecycle. The sessions focus on integrating security tools for development teams, managing risks in AI development workflows, and implementing runtime governance for agents deployed in production environments. Presentations include techniques for securing AI coding workflows, detecting tool poisoning and prompt injection risks, and conducting AI-focused red teaming. The event also explores the limitations of relying solely on policy-as-prompt approaches, detailing practical runtime governance strategies such as PII interception, tool scope limiting, and audit isolation to ensure secure and compliant AI agent operations.