BlackRock Is Spending $100 Million to Train the Trades Workers Building AI Infrastructure

The world's largest asset manager has identified a constraint on the AI buildout that no amount of GPU procurement can solve: there aren't enough people to wire the data centers.

BlackRock's Future Builders initiative, announced in March through The BlackRock Foundation, will spend $100 million over five years to train and place 50,000 skilled trades workers. The trades in question are electricians, welders, plumbers, HVAC technicians and wiremen, and the stated goal is to unblock the energy and data center infrastructure the AI economy now depends on.

In BlackRock's own words

The firm is not being subtle about the connection to AI. Chairman and CEO Larry Fink put it this way in the launch announcement: "Capital alone is not enough – people are central to building our nation's future." He tied the program to an estimated $10 trillion in U.S. infrastructure investment needed by 2033 for energy, digital and AI systems.

Fink has been blunter elsewhere. At the CERAWeek energy conference, he reportedly warned Trump administration officials that "we're going to run out of electricians" needed for AI data center construction.

John Kelly, BlackRock's Global Head of Corporate Affairs, made the same point in a July 4 interview with the New York Post: "We're investing in both the projects but also the people." Kelly told the Post that the target professions face dramatic shortages just as the U.S. races to become energy resilient. He also stressed that BlackRock has no plans to run training itself. The firm is funding organizations that already have track records and helping them scale.

Why an asset manager cares about apprenticeships

There is balance-sheet logic underneath the philanthropy. BlackRock has become one of the central financial engines of the AI infrastructure wave. It has backed Meta's roughly $27 billion Hyperion data center in Louisiana and led the investor group that acquired Aligned Data Centers at a $40 billion valuation. If those projects stall for lack of tradespeople, the investments stall with them.

The labor numbers back up the concern. The U.S. needs an estimated 300,000 new electricians over the next decade to meet AI data center demand alone, while roughly 200,000 current electricians are approaching retirement. Employment in infrastructure-related trades is projected to grow more than 5% over the next decade, well above the 3% national average. Wages reflect the squeeze. Average hourly pay in construction hit $40.55 in January 2026, and weekly earnings are running 25% above the private-sector average.

How the money moves

Per BlackRock's program page, the $100 million flows as grant capital to nonprofit and workforce development partners across multiple states. Funding covers pre-apprenticeship access, training completion and licensure, plus financial education and digital savings tools built into the programs. A $25 million tranche is being awarded through an open RFP, with the first application window closing July 10, 2026 and grantees to be announced in the fall.

The first grants have already landed. In May, BlackRock announced a $30 million commitment in Texas to train more than 12,000 Texans for electrical careers through partners including Texas State Technical College and Dallas College. Texas sits at the center of the country's data center and grid expansion, so the choice of first state was no accident.

BlackRock is also pulling in a corporate coalition, with Walmart, Home Depot, Carhartt, Google and Meta among the partners, per the New York Post. Alphabet president and CIO Ruth Porat endorsed the initiative publicly and pointed to Google's own electrical worker training program as a template for this kind of public-private workforce investment.

AI anxiety meets AI construction

The initiative lands in the middle of a national argument about whether AI destroys jobs or creates them. Future Builders takes a clear side: the next phase of AI-driven growth runs through human workers with hard skills, not around them. Labor leaders at BlackRock's Infrastructure Summit went further and framed it geopolitically. Building trades union president Sean McGarvey argued the U.S. is in "a cold war with China" over AI and data infrastructure, and said labor needs to be brought into project timelines earlier.

There is an irony here. The industry most associated with automating white-collar work is now the biggest force driving demand for hands-on labor that can't be automated. The compute layer of the AI economy is code. The physical layer is conduit, rebar and 400-amp panels, and for the next decade the people who can install them may be the scarcest resource in the stack.