Why is the AI data center boom running out of people, not just power?

The AI infrastructure story has largely been told through chips, capital, and grid capacity. Uptime Institute's 16th Annual Global Data Center Survey, drawing on responses from more than 1,600 operators worldwide, makes clear that's an incomplete picture. For the first time in the survey's history, meeting AI infrastructure staffing needs entered the industry's top six concerns. Over half of respondents report difficulty finding qualified staff.

We put the question to our own network: where's the talent gap hitting hardest? The answer was decisive. 55% pointed to skilled trades and technical operations, well ahead of engineering and technical design at 18%, power and grid interconnection at 16%, and leadership and project delivery roles at 11%.

That result reframes the conversation. This isn't primarily a shortage of senior engineers or executives. It's a shortage of the hands-on, technical operations workforce that keeps facilities running day to day, and it's the layer of the workforce that gets the least attention in industry commentary.

Why this is different from previous talent shortages

Data centre staffing pressure isn't new. What's new is the compounding effect. Rising rack density is forcing operators into phased builds rather than single large deployments, which multiplies the number of projects requiring skilled technical oversight simultaneously. At the same time, the shift toward AI-specific infrastructure demands expertise that didn't exist as a discrete discipline five years ago: high-density cooling, power procurement strategy, and grid interconnection negotiation, layered on top of traditional data centre engineering and operations.

Our poll suggests the pressure is felt most acutely at the technical operations level rather than at the top of the org chart. That tracks with what the wider market has been signalling: it's far easier to hire a project director than it is to find enough qualified electricians, HVAC technicians, and controls specialists to actually staff a facility once it's built.

The knock-on effect on delivery timelines

This matters because staffing gaps don't just slow individual hires; they slow entire projects. When operators can't resource skilled trades fast enough, project timelines slip regardless of how much capital or land is available. In an industry already navigating community opposition and regulatory scrutiny over grid access, an unresourced project is a vulnerable one, and our results suggest that vulnerability sits closer to the construction and commissioning floor than most conversations acknowledge.

What forward-thinking operators are doing differently

The organisations managing this best aren't waiting for the talent market to catch up. They're building relationships with specialist recruitment partners earlier in project planning, treating workforce strategy, including skilled trades pipelines, as part of the capital planning process rather than an afterthought once construction starts, and taking a genuinely global view of where the right expertise sits.

Why this matters now

The AI data centre race has been framed as a race for compute. Our poll results suggest it's just as much a race for tradespeople and technical operators, the workforce least discussed and, evidently, hardest to find. The operators who solve for this layer of talent early won't just build faster. They'll build with far less risk.

Spencer Ogden specialises in connecting energy and infrastructure businesses with the skilled trades, technical, and leadership talent driving the AI data center build-out. Get in touch to strengthen your pipeline before the gap widens further.