AI Agents for Data Center Operators: Uptime's 2026 Survey Shows Trust Dropping Even as Adoption Rises

The Data Center Is Under More AI Pressure Than Ever, and Operators Are More Cautious About It

For the first time in the history of Uptime Institute's Annual Global Data Center Survey, third-party facilities — colocation, cloud, hosting, and SaaS providers — now handle more enterprise IT workload than corporate-owned data centers: 46% versus 44%, according to the 16th Annual Global Data Center Survey, which collected responses from more than 1,600 participants worldwide, including over 800 data center owners and operators, between April and May 2026.

Uptime's own research director, Andy Lawrence, put it plainly during a webinar on the findings: "All of these are linked, and they're all linked through AI... The pressures it's putting on the entire supply chain, the cost base, the availability of power — all of these are being driven by this massive surge in demand for AI infrastructure."

That surge is showing up everywhere in the survey's numbers. Average modal rack density surpassed 11 kW for the first time — Uptime's research director Daniel Bizo called it a "psychological milestone" — driven by a growing share of AI-focused facilities running above 30 kW. Nearly one-quarter of operators now report at least some racks exceeding that threshold, up from 19% a year earlier.

Trust in AI Is Falling, Not Rising, for the Riskiest Tasks

Here's the part that should reframe how anyone selling AI agents into this market thinks about the buyer: as AI's footprint in data centers grows, operators are trusting it less for the tasks closest to actually running the facility.

The share of respondents who believe AI improves facility efficiency slipped to 52%, down from 58% the year before. Confidence that AI reduces human error fell 11 percentage points, to 40%. And the numbers get starker the closer AI gets to direct control: only 31% of operators trust AI to control equipment settings, and just 16% are comfortable with AI making automated configuration changes.

"This drop in trust... suggests that a lot of operators are nowhere near the point of handing over a lot of control tasks to AI," said Jacquie Davis, an Uptime research analyst, on the same webinar. "It speaks to the importance of human in the loop."

That's not operators rejecting AI — trust is highest, and rising, for lower-risk applications like operational analytics and predictive maintenance. It's a market that has sorted itself into two very different tiers of appetite: AI that watches and flags is welcome; AI that touches equipment is not, yet.

The Staffing Crunch Is the Other Half of the Story

The survey's other headline finding is a talent shortage that's getting worse, not better, after several years of relative stability. More than half of respondents (up from 46% the year before) reported difficulty finding qualified candidates, and 28% said they'd lost employees to competing operators. Electrical specialists and junior operations staff were the hardest roles to fill — exactly the skill sets that AI-driven capacity planning and higher-density racks are putting more pressure on.

Cost, meanwhile, eased slightly as the single dominant worry (down to 38% "very concerned," from 44%) — but only because a wider set of AI-driven problems crowded in behind it. Concern about forecasting future capacity requirements climbed to 76%.

Outages tell a similar story of AI raising the stakes without necessarily raising the risk of individual incidents: for the sixth straight year, fewer operators reported an impactful outage (47%, down from 50%), but the ones that did happen got dramatically more expensive — 71% of operators who had an outage said their worst incident cost at least $100,000, up sharply from 57% the year before. Power failures remain the leading cause, at 56% of incidents, and Uptime's chief technical officer Chris Brown flagged that cooling-related failures are likely to grow as AI-driven rack densities keep climbing.

What This Means for Where AI Agents Actually Fit Right Now

Put the trust numbers next to the staffing numbers and a clear shape emerges: data center operators don't have spare headcount to throw at AI-driven complexity, but they also aren't ready to hand AI the keys to equipment control. The opening for AI agents in this market isn't "autonomous facility operator" — it's the analytics, monitoring, and predictive-maintenance layer where trust is already highest, freeing up the electrical specialists and senior operators who are hardest to hire and hardest to keep.

That's a narrower, more specific use case than the "AI runs your data center" pitch some vendors lead with — and, per Uptime's own numbers, it's the one operators are actually willing to buy into today.