Securing AI agents: Control at the Point of Action

AI agents are reaching production systems on credentials built for humans. Authentication confirms who an agent is. Governance reviews what it is. Neither is designed to evaluate what it does after that. 

As federal, state, and local agencies begin deploying AI agents into mission-critical and administrative workflows, the credential and governance models built for human users don't extend to machine identities — creating a gap this session addresses directly. 

In this webinar, Darrel Lewis maps the agentic AI security market, explains what runtime authorization is and how the Delinea and StrongDM gateway delivers it, and answers the questions customers ask most. 

During this webinar, attendees can expect to learn:

• Where Delinea sits in the agentic AI security market: five categories, which problems belong to which vendor, and why no single vendor covers all five
• Why visibility and control answer different questions, and why this session concentrates on control
• What an AI agent is, and the three places agents get deployed: laptops, cloud and on-premise infrastructure, and inside SaaS platforms
• Why controlling what an agent does is a different problem from confirming who it is, and what runtime authorization adds