AI in ITSM hits 61% adoption, but most teams are only scratching the surface

Sixty-one percent of IT organizations are now using AI in their service management processes. That’s a genuine milestone, and it signals that AI in ITSM has moved from “something to explore” to something most teams are actively doing.

But adoption is not the same as impact. And right now, the gap between having AI and getting real value from it is wider than most IT leaders realize.

Most AI use is still at the guidance level

Our research surveyed 718 IT professionals on how AI is being used across their organizations. The most common applications are technician knowledge support, end-user self-service, and ticket triage. In other words, AI is primarily being used to help people find answers faster and route work more efficiently.

These are genuinely useful outcomes. But they represent the lowest level of what AI can do in a service management context. When we look at maturity levels, 42% of organizations use AI solely for guidance and Q&A. A further 16% have moved into recommendations and light automation. Only 3.3% have reached full autonomous execution, where AI handles workflows and resolves issues end-to-end without human intervention.

The good news is that users actually like it

One finding that often surprises IT leaders is that end-user satisfaction with AI is high. 75% of respondents said their users were either satisfied or very satisfied with the improvements AI has made to their service experience. Only 5.5% reported being dissatisfied users.

That matters because user trust is often the hardest part of an AI rollout. The data suggests that where AI has been introduced thoughtfully, people are embracing it. And when users have good experiences with AI-powered self-service, they use it more, reducing ticket volume and freeing up technician time for higher-value work. The conditions for that cycle are already in place for most adopters.

A third of AI users can’t tell you if it’s working

Here’s where it gets uncomfortable. Of the organizations using AI-powered self-service, 33% are not currently measuring ticket reduction. 36% are unsure how AI has affected their average resolution time.

For IT leaders, that’s a significant problem. Without measurement, you can’t optimize. You can’t identify where AI is underperforming. And you can’t make a credible case for further investment in the business.

The organizations getting the most out of AI are treating it like any other operational system: setting metrics before rollout, reviewing them regularly, and adjusting based on the data. The measurement gap isn’t a technology problem. It’s a discipline one.

Where the industry is heading next

The top strategic priority cited by IT leaders for 2026 is proactively predicting and resolving issues before they affect users, selected by nearly half of respondents. That ambition sits well above where most organizations are today, but it’s a clear signal of where investment is going.

Alongside that, 45% want to reduce ticket volumes through AI-powered self-service, and 41% want better AI tools to directly support their technicians. These three goals point in the same direction: moving AI from a support layer into something that actively drives how the service desk operates.

The distance between where most teams are now and where the leaders will be in two years is real. But it’s also a known path. The organizations closing that gap are the ones who’ve moved beyond adoption and started asking harder questions about outcomes.

Download the State of Service Management 2026 report for the full picture on AI adoption, maturity levels, and what separates early adopters from the rest.