Securing AI Agents: Identity Lifecycle, Least Privilege & Continuous Monitoring

AI agents are multiplying fast, and many orgs are unprepared to secure their identities, access, and permissions at scale. In this interview at RSAC 2026, Frank Vukovits and Matt Lawson (Protiviti) break down how to manage agentic identities with identity lifecycle management, least privilege, discovery, and continuous monitoring.  Key takeaways:  1. Start with identity lifecycle and provisioning for AI agents—not just after-the-fact discovery.  2. Treat AI agents as identities (often with multiple entitlements across systems), not “just another service account.”  3. Anchor controls in fundamentals: least privilege, governance, and repeatable processes for humans + machines + agents.  4. Expect massive scale and velocity (24/7 agents) and plan for continuous monitoring—often using AI to secure AI.  5. Reduce tool sprawl: consolidate into platforms/control planes to improve visibility, reporting, and operations.  Want more cybersecurity insights? Discover more: https://delinea.com #aiagent #aigovernance #identitysecurity