While the AI boom is widely understood as a contest for elite engineers and pricey GPUs, behind the headlines a related struggle is unfolding in the skilled trades. As tech companies pour vast sums into building data centers, a shortage of electricians, technicians and machine specialists is emerging as a critical constraint on growth.
Data centers, with their complexes of servers, cooling systems, and power equipment, form the physical backbone of AI systems, but they cannot be built without a large group of skilled trade workers that has been declining for years.
Industry estimates suggest tens of thousands of additional electricians will be needed annually over the next decade, with even larger deficits projected across construction roles. At the same time, a significant portion of the existing workforce is nearing retirement, reducing the pipeline of experienced workers just as demand accelerates.
Potentially Hefty Salaries
Skilled tradespeople working on data centers are seeing substantial wage increases. Depending on location, data center electricians typically earn between about $60,000 and $93,000 annually, though many skilled workers are making over $100,000 a year, along with job perks such as health insurance and a pension. Reports of $150,000 and even higher for a top expert are not uncommon.
The demanding nature of these projects, often tied to strict deadlines, means that extended hours are common, boosting overall earnings and sometimes allowing longer breaks between assignments.
The nature of these roles is evolving as tech changes. The integration of advanced cooling technologies, automation systems, and high-density power infrastructure is creating hybrid job titles that blend classic trade skills with next-gen technical expertise. The line between trade workers and IT engineers can get blurry as they work together in high tech environments
Despite rising pay, expanding the workforce presents real challenges. Training skilled tradespeople takes years, requiring apprenticeship programs and classroom instruction. Data center projects, in contrast, operate on tight timelines and allow limited margin for error, making it problematic to incorporate less experienced workers.
Filling the Gaps
To address the shortage, companies and industry groups are investing in training initiatives. Programs aimed at expanding apprenticeship capacity and updating skills are gaining ground, sometimes supported by partnerships between tech firms and educational institutions. But the challenge is these are long-term solutions to a problem that is slowing data center buildout today.
For younger workers, astronomical college costs and uncertain white-collar job prospects have made apprenticeship-based careers more attractive. Enrollment and interest in trade programs has risen sharply in recent years, with applications for electrical apprenticeships climbing more than 70% between 2022 and 2024, though still not enough to meet demand.
Unlike many tech roles, trade work requires a physical presence. When a new data center is announced, it can quickly exhaust local labor markets, forcing companies to import workers or offer additional incentives. In some cases, workers face long commutes or temporary relocation to meet project needs.
Adding long term confusion to the issue: despite all of the rosy predictions for job growth, it is not certain to what extent there will be continued demand for tradespeople. Data centers require large crews during construction but far fewer workers once operational. While ongoing maintenance and periodic upgrades will continue to generate jobs, automation may replace some of these jobs.
For now, however, the imbalance is glaring. It is perhaps ironic that the expansion of something as abstract as AI is limited by the lack of staffers who work with their hands. In any case, the growth of AI data centers is encountering a real-world limit to its growth.
