Episode 160: Redefining Data Security for AI Agents

As AI driven systems increasingly automate decision making, organizations face new challenges in securing privleged access. This session explores how policy driven, ephemeral credentials and context-aware authorization can meet compliance demands while supportin rapid automation. Join us to learn how to modernize PAM for machine driven environments.

#Speakers

• Host: Mike Schwartz, Founder/CEO Gluu
• Guest: Justin McCarthy, Co-Founder & CTO at StrongDM
• Guest: Amol Kabe, Chief Product Officer at StrongDM

#Homework

• Slide Deck

#Takeaways

• AI is a new privileged surface area, especially if your workforce naively over-permissions agents to data and systems.
• Mike agrees only with the "AM" part of "put IAM at the center of AI governance".
• "Measure usage, safety, and resilience"? Why not "Risk, Transparency and Accountability"? And then how? Perhaps the metrics could be discussed and defined at a new "GovOps" OpenSSF WG?
• "Design with Guardrails"-- Agreed, but how? Do we need a "Governor Module" ala Apple TV series "Murderbot"? See Medium article: https://gluu.co/murderbot