IT Workers Want AI Autonomy With Human Oversight, Study Finds

IT Workers Want AI Autonomy With Human Oversight, Study Finds

A significant gap has emerged between enterprise technology roadmaps and employee comfort levels with artificial intelligence autonomy. While 40% of IT leaders anticipate digital workplace services will operate autonomously by 2030, 61% of employees believe human oversight remains essential, according to research from TeamViewer.

The findings, drawn from a survey of 4,200 employees across multiple countries and industries, reveal a workforce simultaneously embracing and questioning AI's expanding role in daily operations.

The verification burden

Seventy-five percent of employees now use AI in their daily work, but adoption hasn't eliminated the need for human judgment. Most of these users spend an average of two hours each week checking AI-generated work. More concerning, 51% report they don't know when to trust AI output and when to verify it—a knowledge gap that undermines efficiency gains.

This uncertainty creates a paradox: organizations deploy AI to save time, but workers must allocate significant hours to quality control.

Conditional acceptance

Employees aren't categorically opposed to autonomous AI systems. Thirty-five percent say they're ready for the shift, but with clear boundaries. More than 70% of this group want the ability to define specific tasks for AI rather than granting blanket authority.

Workers identified three conditions that would make autonomous AI more acceptable: stronger protections (39%), access limits (36%), and visibility into AI activity (33%). Another 36% said they would accept autonomy if they received notifications before AI made major changes, while 31% want rollback options.

"Organizations will not succeed by giving AI unlimited authority," said Oliver Steil, TeamViewer's CEO. "They will succeed by building systems that operate safely, transparently and accountably within clearly defined guardrails. Trusted autonomy, not unchecked automation, is what will unlock the next generation of productivity."

IT departments as the testing ground

IT leaders themselves are navigating this transition, with 26% reporting their departments already operate with partial autonomy. The research focused specifically on IT because these departments handle structured, repetitive, and measurable work—characteristics that make them natural candidates for autonomous systems.

Ninety-nine percent of IT leaders said they would delegate at least one task to AI. The most common candidates include system updates (61%), device fixes (53%), and device performance adjustments (49%).

"Autonomy will scale when organizations can prove [AI] works in specific, trusted use cases, then expand it with confidence," said Mei Dent, TeamViewer's chief product and technology officer.

Why it matters

The research exposes a critical implementation challenge for enterprises racing to deploy AI: technical capability is outpacing organizational trust. Companies investing in autonomous systems may face adoption resistance unless they build transparency, control mechanisms, and clear accountability into their deployments. The two-hour weekly verification burden suggests current AI implementations haven't yet achieved the reliability threshold needed for true autonomy—a gap that could slow ROI and create employee frustration.

The findings were first reported by No Jitter, based on TeamViewer's research report.