From AI Pilot to Production: Putting Identity Controls to Work [Intermediate]

Moving an AI agent into production means deciding what it can access, whose authority it uses and how that access ends. Using a worked enterprise scenario, Michael Lang explores how to connect discovery, accountable ownership, least-privilege access and lifecycle governance across AI agents and the non-human identities they depend on. The session follows an agent from onboarding through a change in responsibilities to retirement, highlighting where controls must adapt. Attendees will leave with a practical checklist for identifying governance gaps and prioritising the controls needed to scale AI with confidence.