Autonomous but Not Controlled
AI Agent Incidents Now Common in Enterprises
Released: 04/20/2026
AI agents are rapidly becoming embedded across enterprise environments. They span cloud platforms, SaaS applications, internal systems, and LLM-driven workflows. As these systems take on more autonomous roles, organizations are shifting from experimentation to governance. They're redefining how they control, monitor, and integrate AI agents into security and risk frameworks.
This survey report, commissioned by Token Security, explores how organizations are currently managing AI agent governance. While most enterprises report strong visibility into their agents, the data reveals widespread shadow AI and frequent security incidents.
The research highlights a shift toward risk- and context-driven policy models. Organizations base their governance decisions on action risk and human authorization rather than static permissions. At the same time, agentic AI lifecycle management is maturing, though weak decommissioning practices are creating long-term risk.
Ultimately, this report shows that AI agent governance is evolving into a connected system spanning visibility, lifecycle controls, policy design, and monitoring. As agents scale, organizations must move toward a cohesive governance model aligned with real-world operational impact.
Key Takeaways:
- The majority of organizations (53%) operate agents autonomously for low-risk tasks with human review for higher-risk actions. Only 13% of organizations report fully autonomous models. Monitoring is also largely periodic (59%), reinforcing a governance model based on checkpoints and escalation.
- While 68% report high confidence in their visibility, 82% have discovered shadow AI agents in the past year. These agents most commonly appear in internal automation environments (51%) and LLM platforms (47%).
- Organizations are improving front-end lifecycle practices. The majority (59%) report clear documentation of agent purpose and 68% conduct permission reviews. However, only 21% have formal decommissioning processes. Just 19% express high confidence that they fully retire their agents.
- Organizations are converging on action risk (63%) and human authorization (53%) as the primary signals for governing agent behavior. Nearly 79% view context-aware controls as important or very important, and 66% report clear guardrails defining agent boundaries.
- AI agent-related incidents are common, with 65% reporting at least one in the past year. These incidents have tangible business impact, including data exposure (61%) and operational disruption (43%). As a result, organizations are prioritizing monitoring (28%), risk management (29%), and permission control (19%).
Download this Resource
Best For:
- Chief Information Security Officers
- Cloud Security Architects & Engineers
- AI/ML Security Specialists
- Risk and Compliance Professionals
- DevSecOps Engineers
About the Sponsor
%20(2).png)
Token Security provides a comprehensive solution for managing and securing non-human identities. Our platform offers complete discovery and visibility for all types of non-human identities across various environments. Using our advanced AI-based engine, we deliver critical security insights including identity ownership, dependency tracking, usage monitoring, detection of malicious behavior, comprehensive risk analysis, and more. This enables security teams to strengthen identity posture management, proactively enhance threat detection responses, streamline lifecycle management, and remediate risks automatically and more effectively.
Website: https://www.token.security/
