Webinar Description
Key Takeaways
- AI agents are disrupting traditional Zero Trust security models designed for human behavior
- Identity provenance is emerging as a critical focus for securing the AI supply chain
- Practical strategies include real-time, task-specific access and short-lived credentials
- Mapping autonomous identities to human stakeholders is essential for accountability
As artificial intelligence continues to reshape enterprise operations, security frameworks are being put to the test. Episode 5, “Why AI agents break Zero Trust,” explores the growing gap between legacy security models and the unpredictable nature of autonomous agents.
Rethinking Zero Trust in the Age of AI
Zero Trust architectures were originally designed around the predictable patterns of human users. These models rely on anticipating user actions and enforcing strict access controls. However, the rise of AI agents—capable of determining their own requirements and operating with fleeting lifespans—has exposed new vulnerabilities. Incidents involving “goal-hijacking” have highlighted that the core risk is not just excessive permissions, but a fundamental breakdown in identity provenance.
Industry Voices: Addressing the AI Security Challenge
Harish Peri of Okta and Chris Hughes from Resilient Cyber bring their expertise to a focused discussion on securing the AI supply chain. Their conversation moves beyond traditional activity monitoring, emphasizing the need to establish clear identity lineage for every autonomous agent operating within an organization.
Operational Strategies for Modern Security
Participants will gain insight into practical approaches for managing AI-driven environments. This includes granting agents real-time, task-specific access rather than defaulting to broad, permanent permissions. The session also covers methods for mapping each autonomous identity to a responsible human stakeholder, ensuring accountability even as agents operate independently. Replacing static credentials with short-lived tokens is presented as a key tactic for reducing risk in transient agent scenarios.
Why This Conversation Matters Now
As organizations accelerate their adoption of AI, the operational challenges of securing dynamic, machine-driven workflows are coming into sharper focus. The discussion in this episode addresses not only the technical aspects of identity management, but also the broader implications for trust, governance, and risk in a rapidly evolving digital landscape.