For years, Meta has been known for hiring some of the world’s best programmers, but its priorities seem to be evolving in the AI era.
The company has announced LevelUp, a free, four-week fiber technician training program developed in partnership with commercial real estate and infrastructure firm CBRE. The initiative is designed to take people with zero prior experience and prepare them for jobs at Meta’s US construction sites. The first cohorts are expected to begin this summer.
The move is a direct response to a skilled labor crisis that is quietly threatening the entire AI infrastructure buildout. Meta currently operates or is building 27 data centers across the US, and the company has committed $125 billion in capital expenditure for 2026 alone — a 73% jump from 2025. That is a staggering sum of money to spend when you don’t have enough workers to actually build the facilities.
A Shortage That Goes Beyond Fiber
The labor crunch isn’t limited to one role. The broader construction and data center industry is running short on electricians, HVAC technicians, mechanical engineers, and now fiber specialists. The US could face a shortfall of 1.9 million manufacturing workers by 2033, and nearly half a million new construction workers will be needed in 2027 alone. Meanwhile, the construction unemployment rate is near record lows, meaning the available pool is already tapped.
Fiber technicians sit at a particularly critical node in this ecosystem. They are responsible for the physical cabling that connects the GPU clusters, servers, and networking gear inside data centers — the kind of work that cannot be done remotely, cannot be automated away, and cannot be skipped. No fiber, no data center.
CBRE’s Pat Lynch has said openly that “finding the workforce needed to build a data center is a real challenge.” His firm is now operationally responsible for the LevelUp program, which suggests CBRE sees workforce development — not just real estate services — as a core part of its value proposition to hyperscaler clients.
Big Tech Moves Into Workforce Development
Meta is not the first hyperscaler to run a training program, but the scale and directness of LevelUp is notable. Microsoft has a Datacenter Academy operating through community college partnerships. Amazon runs data center apprenticeships. Google funds local training nonprofits.
What distinguishes Meta’s approach is speed. Four weeks, no experience required, with a direct pipeline to jobs on Meta’s own contractor network. The program targets anyone from recent high school graduates to mid-career workers looking for a change — a deliberate attempt to widen the talent funnel beyond the usual suspects.
Dina Powell McCormick, Meta’s President and Vice Chairman, framed it in explicitly nationalistic terms: “The future of the AI revolution depends on a highly skilled US workforce — one that rises to the challenge of building and maintaining the complex systems that power innovation.”
That framing isn’t accidental. In the current political environment, with tariffs disrupting supply chains and immigration policy squeezing the construction labor pool, being seen as investing in American blue-collar workers carries real strategic value alongside its operational one.
The Physical Foundation of AI
The datacenter construction boom has become one of the defining economic stories of the mid-2020s. Since ChatGPT’s launch in late 2022, spending on data center construction in the US has tripled, and for the first time in recorded history, it has surpassed spending on office construction. The hyperscalers — Meta, Microsoft, Google, Amazon — are collectively on track to spend over $650 billion on AI infrastructure in 2026.
But capital is only one constraint. As the CEO of recruitment firm Randstad put it, “the real constraint on global tech growth isn’t solely related to a shortage of microchips, energy, or capital; it is the severe scarcity of the specialized talent required to build it.” A data center under construction can require 50 to 100 electricians simultaneously, with peak crews often exceeding 1,500 workers across all trades.
The LevelUp program won’t close that gap on its own. But it signals something important about how the biggest players in AI are starting to think about the problem. The intelligence that powers modern AI may run in the cloud, but it is built by people in hard hats. Meta is now betting that training those people itself is faster than waiting for the market to produce them.
Since 2010, Meta’s data center projects have supported more than 30,000 skilled trade jobs during construction and over 5,000 permanent operational roles. LevelUp is an attempt to ensure those numbers can keep growing — because without the workers, the $125 billion doesn’t build anything.