AI data center boom stalls as skilled labor shortages hit construction

The AI industry’s explosive growth has outpaced not just hardware and energy supplies, but also the human workforce required to assemble the facilities that power it. According to Saint-Gobain CEO Benoit Bazin, the data center construction sector is now grappling with a labor crisis that threatens to slow the expansion of critical infrastructure. Speaking during a Bloomberg Television interview, Bazin acknowledged that demand for new data centers remains robust, but he warned that a shortage of skilled labor is becoming a major bottleneck across North America and parts of Europe.

Bazin’s remarks highlight a growing concern within the industry: constructing a modern AI data center requires far more than capital investment. Unlike traditional commercial buildings, these facilities demand highly specialized workers, including electricians, high-voltage technicians, fiber-optic installers, HVAC specialists, and controls engineers. Many of these roles require years of training, making it difficult to rapidly expand the labor pool in response to surging demand. "You can’t just hire any construction crew for a data center," said Bazin. "The skill set needed is entirely different, and the training period is substantial."

Power remains the primary constraint, but labor is the next hurdle

While power infrastructure and utility capacity remain the most immediate challenges for AI data center deployment—with even decommissioned nuclear plants like Three Mile Island being reconsidered—industry leaders now point to labor as the next critical bottleneck. Hyperscalers such as Amazon, Microsoft, Google, Meta, and Oracle have committed hundreds of billions of dollars to new facilities, yet the physical construction of these sites depends on a workforce that is not keeping pace with investment.

The shortage extends beyond construction workers. High demand for electricians, in particular, has created ripple effects across other sectors. In Texas, where AI data centers are proliferating, contractors building residential housing have reported delays as they struggle to compete with the wages and budgets offered by hyperscaler-backed projects. "The competition for skilled labor is no longer local—it’s regional, even national," explained an industry analyst. "Housing developments are being delayed by months, not weeks, because of the labor squeeze."

Workforce development efforts emerge as a partial solution

Recognizing the urgency, some technology companies are taking proactive steps to expand the talent pipeline. Meta, for instance, recently partnered with CBRE to launch a training initiative aimed at increasing the number of fiber-optic technicians and other specialized workers. The program, which focuses on upskilling existing laborers and attracting new entrants, reflects growing fears that labor shortages could eventually delay or derail AI infrastructure projects.

However, workforce development alone may not be enough to address the scale of the challenge. The U.S. Bureau of Labor Statistics projects that the need for electricians alone will grow by 7% over the next decade, a rate that may prove insufficient given the AI industry’s breakneck expansion. "We’re in a race against time," said a senior executive at a major construction firm. "The training programs are a start, but we need systemic changes to attract workers into these trades."

Community opposition adds another layer of complexity

Beyond labor and power constraints, data center projects are increasingly facing public pushback. Residents in communities slated for new facilities have raised concerns about rising electricity and water consumption, noise pollution, and the broader environmental impact of large-scale data centers. In St. Paul, Minnesota, a rally drew hundreds of citizens opposing a proposed data center, mirroring similar opposition movements in Texas and other states.

While opposition has not yet halted major projects, it has contributed to delays and increased scrutiny. Lawmakers in some regions are now considering moratoriums or stricter regulations, adding another variable to an already complex deployment landscape. "The industry can’t ignore the social license to operate," noted a policy analyst. "Public sentiment is becoming as critical as technical feasibility."

The road ahead: balancing speed with sustainability

The AI infrastructure boom shows no signs of slowing, with hyperscalers continuing to announce new projects at an unprecedented pace. Yet the combination of labor shortages, power constraints, and community resistance suggests that the path forward will be far from smooth. Industry leaders emphasize the need for collaboration—between companies, governments, and educational institutions—to ensure that the workforce can scale alongside the technology.

For now, the race to build the AI-powered future is as much about people as it is about silicon and servers. The question remains: Can the labor ecosystem evolve quickly enough to keep pace with the industry’s ambitions?