Goldman Sachs says America’s energy worker shortage is so dire, it will need AI-powered robots to fill the gap

Around 41% of the current construction workforce is projected to retire by 2031, according to the National Center for Construction Education and Research (NCCER). At the same time, Goldman Sachs Global Investment Research projects that meeting rising U.S. electricity demand could require about half a million additional workers across the power and grid industry by 2030.

The shortage has become so acute that Goldman Sachs says part of the solution may come from the very technology helping drive the demand in the first place: AI-powered robots.

“We ​need those who have [General Educational Development qualifications] to PhDs, and really anything in between, to support the historic opportunities for building the energy workforce right now,” Anirban Basu, chief economist at Associated Builders and Contractors, told Reuters Events.

Breaking into the field

For workers looking to break into the energy industry, some of the jobs in highest demand can come with six-figure salaries.

Electrical power-line installers and repairers, for example, earned a median salary of $92,560 in 2024, according to the Bureau of Labor Statistics (BLS). Those working directly for utilities earned a median $102,050. The BLS expects about 10,700 openings for these workers each year through 2034, in part because existing workers are leaving the field or retiring.

Getting into the field can also look very different from taking on the cost of a traditional college degree. Lineworkers typically need a high school diploma rather than a four-year degree, followed by technical instruction and extensive on-the-job training. Many enter through registered apprenticeships, which allow workers to earn a paycheck, which grows with their skills while they train.

The problem is that there are not enough people making their way through that pipeline.

Goldman Sachs says there were roughly 45,000 active energy-related apprenticeships in the U.S. in 2024. That pace would need to climb to around 65,000 a year to meet the expected need for workers in transmission and distribution alone.

And the AI boom is giving companies little time to close it.

Data centers are becoming enormous consumers of electricity. Berkeley Lab estimates they could account for about 11.8% of all U.S. electricity use by 2030, with some scenarios putting that share as high as 15.3%.

“Power is a critical bottleneck — but increasingly, the requisite labor presents a structural constraint of its own,” the report explains. “The technical workforce that constructs, wires, cools, and secures this infrastructure is in acute demand, and training cannot happen at the pace capital is being committed.”

Moneywise reached out to Goldman Sachs for additional comment on the report but did not hear back before publication.

Can robots fill the labor gap?

If America cannot train enough workers quickly enough, Goldman Sachs believes technology could eventually help pick up some of the slack.

The firm says growing interest in “physical AI,” including humanoid robots, is being driven in part by the need for labor. Goldman projects the number of humanoid robots on the market could climb from roughly 20,000 in 2025 to 1.4 million by 2035.