Data Centers Face a Retirement Wave Just as AI Demand Climbs
A survey by Onnec of senior data center leaders in the UK, Ireland and Nordic countries expects about a quarter of staff to retire within ten years.
According to TechRadar, a survey of 300 senior decision-makers in the data center industry suggests that a large share of the sector's workforce could leave through retirement in the next five to 10 years. The research, carried out by Onnec, covers operators in the UK, Ireland and the Nordic countries and projects that 23 percent of staff are expected to retire over the coming decade.
TechRadar reports that the research places the replacement need in a range of 18 to 25 percent of roles. The outlet also says 10 percent of the businesses surveyed expect more than 40 percent of their employees to retire in that period, a figure that points to very uneven exposure across companies.
The timing is notable because the industry is expanding, which TechRadar links to growing demand for artificial intelligence. More facilities and more capacity mean more roles to fill at the same moment that experienced workers are approaching retirement age.
The findings have led to discussion of apprenticeships and mentoring as possible responses, according to the outlet. The coverage presents these as options under consideration rather than announced plans, and it does not detail which specific roles are most affected or how individual companies intend to respond.
Background
Data centers are buildings that house computing equipment, along with the power, cooling and network systems needed to keep it running. Staffing them takes a mix of skills that rarely overlap neatly. Electricians and mechanical technicians keep power and cooling systems working around the clock, facilities managers coordinate maintenance and safety, and network and systems engineers handle the digital side. Many of these roles are learned on the job over years, which makes them slow to replace when experienced people leave.
Workforce aging is a familiar concern across many skilled trades and engineering fields in developed economies. When older workers retire, they take with them informal knowledge about specific equipment and failure patterns, which written manuals capture only partly. Employers often respond through a combination of apprenticeships, which pair classroom training with paid work, and mentoring, where experienced staff guide newer colleagues over an extended period.
The sector has grown quickly as cloud computing and streaming expanded, and more recently as companies build capacity for artificial intelligence training and inference. Those workloads draw large amounts of electricity and generate heat, so facilities need careful engineering and steady maintenance. The UK, Ireland and the Nordic countries have each attracted investment for reasons that include power supply, climate and regulatory environments, which means labor pressures in these markets can be shaped by local competition for the same pool of technicians.
Why it matters
A workforce shortage in data centers is not only a hiring problem. These facilities support banks, hospitals, government services and everyday online tools, so staffing gaps can affect reliability and the pace at which new capacity comes online. If operators cannot find qualified technicians, construction schedules and maintenance routines may slip, and the costs of recruiting and retaining people may rise. For workers, the situation could create opportunities for people entering from other trades or from training programs, though pay and career paths will depend on how employers respond. The findings also raise a planning question for policymakers: whether national training systems can supply enough skilled labor for a sector that is growing fast. Companies that invest early in structured training may gain an advantage, while those that rely on poaching staff from competitors could find the market increasingly expensive. The survey is a useful signal, but it reflects senior leaders' expectations, so the gap between forecast and reality will be important to track.
What to watch
Watch whether data center operators announce concrete apprenticeship or mentoring programs, and how many places they fund. Look for moves by training providers and colleges in the UK, Ireland and the Nordic countries to add courses in facilities engineering and data center operations. Pay trends for technicians and electricians will show whether the expected shortage is tightening the labor market. It will also be worth seeing whether companies publish retention data, and whether AI-driven demand for new facilities continues at its current pace, since that will shape how many roles need filling.
Viewpoints
Commentary from our columns: three angles on this story.
The Ledger the business angle
The retirement numbers in this survey are a balance-sheet problem before they are a people story. Data center construction is capital-intensive, and every month a facility waits for qualified staff is a month of delayed revenue. Operators who have spent heavily on land, power and equipment have a clear incentive to treat staffing as a core input rather than an afterthought. What is less clear is who pays for training. Apprenticeships cost money up front, and trained workers may leave for a competitor. That creates a familiar dilemma: each company would prefer that others train the workforce while it simply hires. Shared training consortiums, co-funded programs or clear retention terms could address this free-rider problem. The business case for investing now looks stronger than waiting. Firms that build pipelines early may face lower recruitment premiums later and will be better positioned if AI demand keeps pushing the sector to expand. The survey gives boards a reason to put workforce planning on the same agenda as power contracts and site selection, where it arguably belongs.
The Builder the engineer's view
From an engineering standpoint, the real risk here is knowledge loss. A facility's cooling loops, power distribution and failover procedures are often understood best by the people who have maintained them for years. When those engineers retire, documentation may not capture everything they knew about how a system behaves under stress. Teams should treat this as a systems problem: write proper runbooks, record incident reviews, and use simulation tools where they help new staff learn safely. Mentoring works best when it is built into shift schedules and project work rather than squeezed in after hours. Apprentices need supervised time on real equipment with clear progression goals. AI infrastructure adds pressure, because higher-density racks and liquid cooling change the skills required. A technician trained on traditional air cooling may need new preparation to work safely with newer designs. Engineers designing the next generation of facilities also have a role. Simpler maintenance, modular components and better monitoring can lower the skill threshold for routine tasks, letting a smaller group of specialists cover more ground. The best answer to a retirement wave is often a system that people who are still learning can run well.
The Optimist the hopeful reading
There is a hopeful reading of this survey. A workforce that must be replaced is also a workforce that can be rebuilt on better terms. Data center jobs have long been hard to see from the outside, and many people do not know the sector needs electricians, technicians and engineers in regions that may lack strong local employment. Expanding apprenticeships could open these careers to people who were never told they were an option, including those changing careers in midlife. Mentoring offers retiring experts a chance to pass on knowledge built over decades, and many experienced workers may welcome a role in shaping the next generation. The fact that industry leaders are discussing training openly, not only hiring, is a sign that the conversation is maturing. Problems like this are solvable when people start planning early. The UK, Ireland and the Nordic countries have strong records in technical education and vocational training to build on. If the sector treats this moment as an opportunity to create stable, skilled jobs, the retirement wave could leave behind a more diverse and better-prepared workforce than the one it replaces.