90% of IT Leaders Plan for Autonomous Operations, But Most Still Require Human Oversight

Nine out of ten IT leaders now plan to use agentic AI to run their IT operations independently, according to Riverbed's global survey of 1,200 business decision-makers, IT leaders, and technical specialists. But nearly as many—77%—say their organizations will not allow AI to make operational decisions without a human in the loop.
The survey, titled The State of Autonomous IT Operations, found that 76% already consider agentic AI critical or very important to their future IT strategy. Investment is rising: 63% of respondents have increased AI spending, and 91% say those investments have met or exceeded expected ROI.
Despite the enthusiasm, execution lags ambition. Only 19% of IT operations are currently automated, and just 8% report enterprise-wide AI-enabled operations. Organizations expect autonomous IT to become a reality within three years, but readiness gaps remain wide.
Data quality and multi-domain visibility are identified as the two foundational blockers. Ninety-six percent of respondents say cross-domain visibility is critical, but only 17% have unified it across networks, applications, endpoints, and cloud—a 79-point gap. Similarly, only 23% rate their data granularity as excellent and 21% say the same of data quality.
Tool sprawl is a further obstacle: 91% agree consolidation reduces friction, and 90% say a unified platform simplifies problem identification.
Key Takeaways
- 90% of organizations plan to pursue human-supervised or highly autonomous IT within two years; only 19% have automated IT today
- 77% require human approval before AI can take operational actions—trust and governance are the primary blockers
- 92% expect AI observability and governance to become a critical new IT domain, yet readiness remains low
Read the full article on Business Wire
Stay in Rhythm
Subscribe for insights that resonate • from strategic leadership to AI-fueled growth. The kind of content that makes your work thrum.
More from Thrum
Additional pieces exploring adjacent ideas
