Governing AI, Agents, and Autonomy Across the Modern Enterprise | SailPoint Navigate 2026

BLUF: Organizations must urgently evolve their identity governance frameworks to address the exponential growth and unique risks of AI agents, moving beyond traditional static controls to dynamic, risk-tiered governance models 2:15.

Key Takeaways:
• AI agents are proliferating rapidly, with surveys indicating they double every quarter and 38% of organizations have over 100 in production, yet 90% have reported AI-related breaches 5:40.
• Traditional non-human identity (NHI) playbooks fail for agents because agents are dynamic, autonomous, and non-deterministic, making ownership, least privilege, and predictable auditing nearly impossible with current tools 11:00.
• Governance for agents requires a shift from static controls to a "closed-loop" model focusing on continuous discovery, dynamic authorization, and real-time observability to manage unpredictable behavior 25:40.
• Not every agent requires full Identity Governance and Administration (IGA); organizations should implement risk tiering to apply appropriate controls based on the agent's potential impact and risk profile 27:50.

Closing Statement: Security leaders must prioritize immediate discovery of existing agents and assess their current governance readiness against these new dynamics to build a business case for necessary infrastructure updates.

Sources:

• 5:40 Statistics on AI agent growth and breach prevalence.
• 11:00 Comparison of traditional NHI management vs. agent challenges.
• 25:40 Introduction of the closed-loop governance framework for agents.
• 27:50 Risk tiering strategy for managing agent governance scope.