JumpCloud Inc. has released a new study, “Enterprise AI Access Is Outpacing Control,” highlighting a growing gap between AI adoption and security governance. The report finds that more than half of organizations now use AI agents capable of modifying systems, permissions, or files. Yet, among organizations using these agents, 59% do not apply the same governance policies to AI agents that they use for human users.
The Enterprise AI Access Is Outpacing Control study of 250 IT leaders across the U.S. and U.K. shows how organizations are managing AI as it gains greater access across the enterprise. The research also found differences in AI auditability across workplace software platforms: just 11% of Google Workspace-primary organizations fall into the lowest auditability tier, compared with 21% of Microsoft 365-primary organizations.
The findings reveal significant gaps in how organizations govern AI access and activity:
- The Identity Gap: Among organizations using AI agents, only 41% have fully integrated agents into the same IAM policies used for human users, leaving 59% without full integration.
- The Proof Gap: 72% of organizations report a considerable gap between what AI is allowed to access and what they can prove it actually accessed.
- AI-Related Incidents: 92% of organizations report experiencing at least one AI-related incident or exposure in the past 12 months. More than one-third (36%) investigated an AI-related concern but were unable to determine what the AI had accessed.
“AI agents are quickly becoming another class of identity in the enterprise. They can access sensitive systems, change records, trigger workflows, and act on behalf of people, but many organizations still aren’t governing them with the same rigor they apply to human users,” said Joel Rennich, senior vice president of advanced technologies & innovation, JumpCloud. “Every agent needs clear ownership, appropriate access, and controls that follow it throughout its lifecycle. As agents become more autonomous, that foundation is what allows organizations to scale AI without losing accountability.”
To solve this challenge, the report details the Agentic IAM Lifecycle framework. This guide helps IT teams discover, register, manage, and govern AI agents alongside human employees. By applying identity rules to AI tools, companies can assign owners, set permissions, and track AI actions start to finish.
To review the Enterprise AI Access Is Outpacing Control study, visit the website here.
Related News:
