Big Tech Is Pouring Billions Into Electrician Training for AI Data Centers — Skilled Trades Have Never Been More Valuable

Meta, Google, and BlackRock have committed more than $215 million to skilled-trades training programs for AI data center construction, with Meta alone spending $115 million on its America's Workforce Academy in Louisiana. OpenAI told the White House its infrastructure plans require a fifth of all US skilled tradespeople. The AI buildout is creating a labor crunch so severe that tech companies are building their own training pipelines for electricians, carpenters, and HVAC technicians, with data center electrical work now commanding $40-$55 per hour.

Big Tech Is Pouring Billions Into Electrician Training for AI Data Centers — Skilled Trades Have Never Been More Valuable Meta, Google, and BlackRock have committed more than $215 million to skilled-trades training programs for AI data center construction, with Meta alone spending $115 million on its America's Workforce Academy in Louisiana. OpenAI told the White House its infrastructure plans require a fifth of all US skilled tradespeople. The AI buildout is creating a labor crunch so severe that tech companies are building their own training pipelines for electricians, carpenters, and HVAC technicians. Data center electrical work now commands $40-$55 per hour, pulling workers away from residential construction. The Numbers Are Staggering OpenAI /glossary/openai , Google, Meta, and BlackRock are recruiting electricians, carpenters, and HVAC technicians by the thousands, and they're offering pay that the construction trades have never seen. According to a Times report published July 28, the demand for skilled tradespeople to build and maintain AI data centers has created a labor crunch so severe that tech companies are now building their own training /glossary/training pipelines. - <<<BOLD Meta:<<<BOLDEND $115 million committed to America's Workforce Academy, training recruits in Louisiana - <<<BOLD Google:<<<BOLDEND $10 million pledged to the Electrical Training Alliance - <<<BOLD BlackRock:<<<BOLDEND $100 million directed to skilled-trades training programs - <<<BOLD OpenAI:<<<BOLDEND Told the White House its infrastructure plans alone need a fifth of all US skilled tradespeople The AI buildout is no longer just about chips and models. It's about concrete, copper, and cooling systems. And the companies racing to build it are running out of people who know how. Why Trades, Why Now A single hyperscale data center requires 30 to 50 megawatts of power capacity — enough to supply a small town. Each facility needs miles of conduit, thousands of electrical terminations, specialized cooling infrastructure, and physical security systems that require licensed tradespeople to install and maintain. Traditional construction labor pipelines can't scale to meet demand. The Electrical Training Alliance, a joint labor-management program, typically trains about 10,000 apprentices per year across the US. Google's $10 million pledge adds capacity, but it's a drop against the estimated 200,000-plus skilled trade roles that the AI data center boom will create over the next three years. Meta's Louisiana-based America's Workforce Academy is the most concrete example. It's a dedicated training facility that fast-tracks electrical and mechanical apprenticeships specifically for data center work. Recruits go through an accelerated curriculum focused on the specific power and cooling systems used in AI infrastructure. Graduates walk into jobs starting at $35-plus per hour with overtime. The Pay Shift Data center electrical work now commands a premium over traditional commercial construction. Union electricians in data center markets like Northern Virginia, Dallas, and Phoenix report hourly rates of $40 to $55, with unlimited overtime during build phases. Non-union shops are matching or exceeding those rates to compete. The Times report highlights that some of this recruiting is pulling tradespeople away from residential and light commercial work, creating ripple effects in housing construction costs. A framer or electrician who can earn $50/hour at a data center has little reason to show up to a $30/hour residential job. What This Means The AI industry is discovering that infrastructure is the bottleneck — not models, not chips, not data. You can have the most capable model in the world and it's worthless if you don't have a trained electrician to connect the racks. For anyone considering a career change, the signal is unmistakable. The trades — especially electrical, HVAC, and carpentry — are now a direct path into the AI economy. No coding required. Key Takeaway The $100 million-plus from BlackRock, Meta, and Google into trades training isn't philanthropy. It's supply-chain hedging. Tech companies are competing with every other industry for a finite pool of skilled labor, and they're building their own pipelines because the market isn't producing enough fast enough. If you can wire a 480-volt three-phase panel, you're more employable in the AI economy than most junior software engineers. Get AI news in your inbox Daily digest of what matters in AI.