GPSEC Greatest Hits:
The Identity of AI Agents: Securing Them from Themselves & Outsider Threats

The Identity of AI Agents: Securing Them from Themselves & Outsider Threats

AI agents aren’t human or machine—they’re something in between. They are actors with privilege and agency, not scripts. Their non-deterministic behavior means they can’t be secured like typical identities. Each must be mapped to a human initiator, leaving no orphaned or rogue agents without accountability. To stay ahead, organizations must treat them as autonomous entities with owners, scope, observability, and instantly revocable access. Join Roy Akerman, VP of Identity Security at Silverfort and founder of Rezonate, who brings MIT data science expertise and national cyber defense experience to reveal why securing agentic AI identities is key to resilience, and explain how to take an identity-first approach to securing AI agents.

Speaker: Roy Akerman, VP of Identity Security, Silverfort

SEE THE FULL SCHEDULE HERE