Organizations Rapidly Reducing Human AI Oversight
A new report reveals that organizations are increasingly deploying AI agents with full autonomy, doubling the attack surface.
A significant shift towards artificial intelligence (AI) autonomy within organizations is underway, according to a new Q3 2026 IT Trends Report released by JumpCloud Inc. The report indicates that over six in 10 organizations now operate AI agents in production environments. However, governance has not kept pace, with organizations adopting less than a third of standard AI governance and security practices.
Key findings show a dramatic decrease in human oversight for AI. The percentage of organizations requiring human review before high-risk AI actions dropped from 40% to 25% in just six months. Concurrently, full autonomy for AI agents without human review more than doubled, rising from 11% to 26%. This trend expands the potential corporate attack surface.
The report also highlights a decline in self-assessed AI maturity, with only 23% of IT leaders describing their organizations as mature in AI deployment, down from 40% six months prior. This is occurring as AI agents gain broader access and delve deeper into critical workflows. Furthermore, non-human identities now outnumber human users in 83% of organizations, yet only 21% have adopted governance controls for them.
"Intelligent, secure IT for the agentic era is not a future state. It is the requirement right now," stated Joel Rennich, senior vice president of product management at JumpCloud. He emphasized the need for organizations to treat every identity, human or not, with rigorous security, management, and governance to scale AI confidently and securely.