AI agents are evolving from passive assistants to autonomous, task-executing entities, capable of acting on behalf of users with minimal oversight. This shift introduces critical security risks, from "vibe coding" to locally hosted AI assistants operating outside IT control.

The panel will explore how these autonomus AI systems create new attack surfaces, complicate identity management, and enable shadow AI, where unsanctioned AI tools operate undetected within organizations.

The discussion will also examine the challenges of non-human identity management, where AI agents act with delegated authority but lack traditional accountability.

This session will dissect the security implications of a world where AI doesn’t just assist, but acts, often with limited human intervention.

Tune in to this session to learn:

• How to implement AI-specific access controls by applying just-in-time (JIT) permissions, behavioral monitoring and least-privilege principles
• Strategies to detect and govern shadow AI agents through network segmentation, AI aware security solutions and policy frameworks to secure unsanctioned AI tools
• How to integrate clear mechanisms to monitor, audit and revoke access for AI systems that autonomously perform tasks