JumpCloud Research Reveals AI Agents are Operating Without Governance in Critical Workflows

jumpcloud-research-reveals-ai-agents-are-operating-without-governance-in-critical-workflows
JumpCloud Research Reveals AI Agents are Operating Without Governance in Critical Workflows

 72% of Organizations Have Agents in Production, but Most Lack Essential Controls

, /PRNewswire/ — JumpCloud Inc. today released its latest research findings, The Agentic IAM Pulse Report: Closing the Governance Gap to Accelerate with AI. The report shows that AI agents are quickly entering key business operations. However, organizational governance is struggling to keep up. While 72% of organizations have AI agents in use, a staggering 92% report serious limits in safely scaling their deployments.

The study highlights a dangerous trend. AI agents are shifting from testing to key tasks like financial reporting and HR provisioning. They now have more freedom and less oversight.

Key Findings Include:

  • The access paradox: 66% of organizations grant AI agents equal or greater system access than human employees. This gap is most acute in business-critical environments, where 38% of agents receive significantly more access than their human counterparts.
  • Declining oversight at scale: Human-in-the-loop (HITL) approvals fall from 48% in testing to 29% in key business deployments. 24% of organizations allow agents to execute high-risk actions with zero human supervision.
  • The identity explosion: 53% of organizations now manage more non-human identities than human employees, with 23% reporting a ratio of six-to-one or higher. This scale makes manual management impossible, necessitating an Agentic IAM solution to autonomously govern these environments where human oversight can no longer keep pace.
  • Fragmented accountability: Only 17% of organizations have a designated security leader accountable for AI agent actions. Most accountability (51%) defaults to IT teams alone for business-critical deployments.
  • The kill switch vacuum: 55% of organizations lack a centralized kill switch to cut AI agent access across all systems.

“AI agent deployment has officially outrun the controls needed to manage it safely,” said Joel Rennich, senior vice president, product management, JumpCloud. “We are seeing agents operate in sensitive workflows with fragmented identities and more access than human employees, yet they are the least supervised group in the enterprise. Because identity is now the only perimeter left for these agents, organizations must move toward a formal governance model that treats every agent as a first-class, governed identity to turn AI from a liability into a sustainable engine for growth.”

Built for the Agentic Era: JumpCloud Agentic IAM

This research shows the need for JumpCloud’s new Agentic IAM solution. Agentic IAM provides the unified control plane necessary to anchor every human, non-human, and autonomous agent to a verified corporate identity. By providing automated guardrails for the entire agentic lifecycle, JumpCloud ensures that AI moves from a shadow risk to a secure competitive advantage.

Download the full report. Gain deeper insights into the four-stage framework of AI Agent governance and the evolving landscape of non-human identity management.

About JumpCloud
JumpCloud® is the AI-powered unified IT management platform designed to secure the modern workforce. By consolidating identity, device, and access management, JumpCloud provides intelligent, secure IT that scales from human users to autonomous AI agents. We help organizations around the globe eliminate complexity and turn AI risk into an optimized advantage, ensuring the right people and agents have secure access to the right resources at all times.

Learn more: jumpcloud.com/
Follow us: Blog | Podcast | X | LinkedIn | YouTube | Resources

Click here to get started with JumpCloud

Contact
For JumpCloud
Josie Caracciolo
[email protected]

SOURCE JumpCloud, Inc.