Demand for new AI data centers is skyrocketing, but you can’t build them yourself.
Big Tech is pouring billions of dollars into building these specialized facilities with four hyperscalers. alphabet, microsoft, metaand Amazonwill spend nearly $700 billion in combined capital investment this year to fund these developments.
Last month, Amazon announced it would spend $12 billion to build a new AI data center in Louisiana. This will create 540 full-time jobs on site, as well as 1,700 roles including electricians, technicians and security professionals.
Last year, Meta invested $27 billion in joint ventures with blue owl capital The giant Hyperion data center in Louisiana is expected to consume more electricity than the city of New Orleans.
While fears about white-collar jobs being replaced by AI are at a fever pitch, the data center boom is creating lucrative opportunities for skilled trade workers.
“The digital revolution requires a large-scale physical infrastructure,” says Sander van’t Noordende, CEO of the world’s largest recruitment agency. Randstadhe told CNBC. “Ultimately, the real constraint to global technology growth doesn’t just involve a lack of microchips, energy, or capital; it’s a severe lack of the specialized talent needed to build it.”
Demand for robotics technicians increased by 107% between 2022 and 2026, according to a global analysis of 50 million job postings published by Randstad on Wednesday. The growth rate for refrigeration (HVAC) systems engineers was 67%, and industrial automation technician vacancies increased by 51%.
Meanwhile, the number of job openings for traditional skilled trade jobs such as construction workers and electricians increased by 27%, according to Randstad analysis.
“Debates around the impact of AI on the labor market are often completely focused on the software side, and specifically whether generative models will displace white-collar jobs. However, an important reality is completely overlooked: AI cannot build its own data centers,” Noordende said in emailed comments.
With approximately 12,000 data centers around the world today, and thousands more planned to be built to house high-performance AI computing power, it is essential to update outdated mechanical, electrical and plumbing systems every four to six years, said Mike Matthews, digital infrastructure lead at professional services firm Marsh.
Matthews noted that as a result of these retrofit requirements, there was a “significant area of unique growth in the workforce” as network engineers, electricians, mechanical engineers, plumbing and heating contractors were deployed to install new liquid cooling systems.
Matthews, himself a fourth-generation plumber, described these roles as “new-collar” jobs, saying traditional white-collar and blue-collar employees would work side-by-side with each other and receive the same recognition.
“This is the first time in the data center space that we’ll have highly compensated, highly skilled workers physically working next to network engineers with college degrees. And we think this will be a great social blend across the data center,” Matthews added.
Skills shortage drives wage growth
“The scarcity premium is kicking in,” Randstad’s Noordende said, adding that posted wages for HVAC engineers have increased by about 10 to 15 percent over the past four years.
“Wage growth for these professionals is increasing significantly as the demand for AI infrastructure outpaces the shrinking labor supply,” Noordende said, adding that six-figure salaries are achievable in this field.
Professional technicians in high-level data center roles often earn a 25% to 30% salary increase, according to staffing firm Kelly Services. The company shared estimates with CNBC based on internal and third-party data. The company said premiums may vary by role and full range salary data for data center positions has not yet been disclosed.
NvidiaJensen Huang, the company’s CEO at the center of the AI data center boom, predicted in January that “six-figure salaries” for workers building AI factories were on the horizon.
The primary factor driving these salary increases is a shortage of trade workers, with the United States facing a potential shortage of 1.9 million manufacturing workers by 2033, according to 2025 data from the National Association of Manufacturers.
Meanwhile, the U.S.-based Associated Builders and Contractors trade group estimates that nearly 500,000 new workers will be needed in 2027, up from the 349,000 needed in 2026.
“[Skills shortages]are a big problem right now, and it’s only going to get worse,” Gary Wojtaszek, CEO of Pure Data Centers, told CNBC.
“The good thing is that AI is not going to replace these jobs. These are really important jobs because someone needs to operate these machines, and this is a big challenge for the entire industry.”
Randstad’s Noordende said companies and governments needed to invest in training programs to overcome skills shortages.
Skills shortages are also leading to cross-industry poaching because there is a lot of overlap in the operational technical skills needed in the energy, defense and technology sectors, said William Self, principal workforce strategist at Mercer, a global workforce consulting firm.
In early March, black rock CEO Larry Fink emphasized that capital alone is not enough to deliver the $10 trillion in investments needed in infrastructure, and launched a $100 million initiative to empower the next generation of trade workers.
“Skill profiles are evolving faster than traditional job descriptions can track,” Mercers-Self said at a virtual press conference on the subject.
“Given the shortage, the companies that win this talent race will be those that invest in both traditional recruiting and non-traditional workforce development. I’m looking at apprenticeship programs, community college partnerships, veteran pipelines, and even internal talent academies to develop companies’ own talent.”
“Hazard pay rate”
However, the growing demand for key AI infrastructure jobs also faces several hurdles, including an aging population and geographic constraints.
According to Noordende, approximately one in four workers around the world is nearing retirement age, and the talent pool is not being fully replenished.
“Unlike software developers who can work remotely, skilled professionals have very low geographic mobility,” he said.
“Equipment technicians and construction workers need to be physically present on site. When a company builds a new AI data center or manufacturing plant, it’s a region-changing event that can quickly deplete the local talent pool.
What is difficult to predict going forward is how much “hazard pay” will be included in paychecks in the future, Mercers-Self said.
Earlier this month, an Amazon Web Services data center in the United Arab Emirates was targeted by an Iranian drone attack. As conflicts in the Middle East unfold, Self added, “I can also imagine that there will be some sort of psychological strain on people who are working in facilities where they know they are likely to be a target or a bad guy. That may further increase the types of compensation that we look at to attract these people to these facilities.”
