Register to build your agenda.
Building Guardrails and Kill Paths for Agentic I&O
Tuesday, December 08, 2026 / 02:30 PM - 03:00 PM PST
About the Session
As agentic AI systems automate IT operations, traditional controls cannot prevent outages and untraceable incidents. We introduce a three-layer architecture: governance, control and assurance, for operationalizing AI safety. Learn how to route all AI actions through a central gateway, enforce strict limits on scope and privilege and implement rapid containment mechanisms like kill switches and privilege revocation. Gain practical steps to make each AI action auditable, explainable and reversible.
Speakers