The Zero Trust Guardrail for Agentic AI

Traditional automation like scripts have static intent. AI agents, however, can expand their actions beyond initial prompts, creating a dynamic attack surface that shatters conventional trust models. This reality demands a shift to a Zero Trust strategy that treats every agent as a privileged, untrusted identity. This session explores a pragmatic approach to this challenge. We will discuss foundational steps like discovering unsanctioned AI and securing their secrets, as well as the need for agent lifecycle management.

We will then look forward, examining how machine-speed authorization—drawing on lessons from using AI to grant real-time, risk-aware access to humans today—can serve as the ultimate guardrail for autonomous systems. Join us to discuss a realistic roadmap for building digital trust in the age of AI.