The AI Paradox: Why IT Teams Are Working Harder Despite Automation Gains

According to the survey of IT professionals worldwide, 84% say AI has met or exceeded ROI expectations, yet 52% report their overall workload has increased since adoption. After an average of about 16 months running AI in their ITSM environments, most teams are still managing the overhead rather than realizing the payoff.

The productivity math that doesn't add up

The time savings are real, on paper. Respondents told SolarWinds that AI saves an average of 3.2 hours per week detecting and flagging issues, 3.0 hours on end-user requests, and 2.9 hours on ticket triage. Those are not trivial figures. But the savings get swallowed by a new category of work that did not exist at the same scale before AI. Eighty-three percent of respondents now spend three or more hours per week just keeping their AI systems running reliably. The math inverts.

The cost surprises compound it. The top unexpected expenses — staff training at 48%, data quality and cleanup at 47%, and ongoing tuning and maintenance at 45% — are recurring, not one-time setup costs. Only 7% of respondents say the cost of AI adoption matched what they planned for. This is what negligence dressed up as innovation looks like in the line items: budgets balloon, labor shifts, and nobody flagged it during procurement.

Reactive by design

The most uncomfortable finding sits in the incident lifecycle data. When asked where AI has had the greatest impact, respondents cited identifying issues before they impact users (31%) and prioritizing and routing issues (23%). Both are fundamentally reactive. Only 19% pointed to preventing issues before they occur as the area of greatest impact. The tools are deployed. The infrastructure to move AI upstream into prevention is not.

This is a maturity gap, not a tooling gap. Organizations have the technology and the budget momentum — 85% report their AI in ITSM budget increased year-over-year, with 36% calling that increase significant. The discipline to use AI proactively is missing.

Where the spend is heading

Agentic workflows, the most proactive AI capability category in the survey, show the highest expected investment growth of any area measured. Read that against the current reactive posture and the picture clarifies: buyers know they are behind on maturity, and they intend to spend their way out.

For enterprise IT leaders, the actionable cut is plain. Track AI operations time as a separate metric from AI productivity gains. Audit the recurring costs — training, data cleanup, tuning — before the next budget cycle. And pressure vendors on prevention, not just identification. The market is rewarding reactive AI heavily and funding proactive AI modestly. That ratio needs to flip.