The companies building artificial intelligence are also pledging to retrain the workers it stands to disrupt — and the timing is pointed. In the span of a few weeks in June, two of the largest U.S. technology firms committed a combined $165 million to worker-training programs, even as the broader sector sheds jobs to fund its AI ambitions.
What's been pledged
Meta launched America's Workforce Academy on June 10 with an initial $115 million investment, Fortune reported. It is a free, five-week skilled-trades program — covering tuition, housing and a training stipend — with a job offer for graduates, starting with pilots in Louisiana, Ohio, Indiana and Texas. The trades it targets are telling: fiber technicians, electricians, welders, plumbers and other roles needed to physically build the AI data centers Meta is racing to construct. The company called it the largest private-sector commitment to skilled trades with a job guarantee in U.S. history.
A day later, Google.org pledged $50 million to help train more than 300,000 skilled-trade workers across more than 20 states, according to Axios, working with labor unions and trade associations and weaving AI tools into apprenticeship programs. Verizon, which announced its largest-ever round of layoffs in late 2025, paired those cuts with a $20 million reskilling fund and its chief executive has urged Fortune 100 peers to set up similar funds — though no other has publicly matched it.
Why now
The pledges land against a wave of AI-related job cuts across the technology industry, where firms are trimming staff while committing enormous sums to AI infrastructure. The training money is, in part, self-interested: Meta's program exists because a shortage of electricians and fiber technicians is a genuine bottleneck for its data-center build-out. The skilled hands that pour foundations and lay fiber are, for now, in short supply.
The skeptics' case
Economists caution that retraining, however well-funded, is not a cure-all. Julian Jacobs, writing for the Brookings Institution, argues retraining should be seen as only one part of a broader response to AI disruption, noting that past federal job-training efforts produced limited measurable gains in employment for displaced workers. Brookings research has also found that millions of U.S. workers facing high AI exposure live in places with little local retraining infrastructure or few alternative employers to absorb them.
There is also a mismatch at the heart of the current pledges. The marquee programs train people to build AI infrastructure — construction and trades jobs — rather than to replace the white-collar and customer-service roles that AI is automating inside offices. A program that produces data-center electricians does little for a displaced administrative worker.
What it means
The corporate commitments are real money and a meaningful signal that the companies driving automation feel some responsibility for its fallout. But at a combined few hundred million dollars, spread across hundreds of thousands of intended trainees, they are modest next to the scale of the disruption — and weighted toward the workers AI's build-out needs rather than the ones it displaces. Whether they amount to a genuine safety net or mostly reputational cover will depend on whether more employers join, and whether the programs reach the people most at risk rather than those already best positioned to benefit.
