AI Agent Security Lab

AI agents are already running in production, and most were deployed without a security review. This programme is built for the practitioners who now have to attack, defend, and govern them. It focuses on hands-on, secure AI agent deployment into a practical, demonstrable capability by building the offensive skills, defensive controls, and governance tooling needed to find real vulnerabilities, close them, and produce the documentation regulators now expect.

Delivered as six live, Python-based sessions, participants will walk through the full attack surface of agentic systems, from prompt injection, tool misuse, and RAG poisoning through sandboxing, trust boundaries, monitoring, and EU AI Act governance. Practitioners execute real attacks, harden the systems they just broke, and build the telemetry and audit trails auditors require, leaving able to secure their own agent architectures and demonstrate compliance readiness with confidence.