The hiring boom you probably did not expect
Talk about AI and jobs usually lands on automation. The part that gets missed: building and running the hardware behind AI takes people. As the likes of Anthropic, Meta, Amazon, and OpenAI scale up, the demand spills into HVAC technicians, construction crews, line workers, pipefitters, and more.
"These are occupations that are becoming increasingly important to the AI economy," said Maria Flynn, president and CEO of Jobs for the Future. She noted data centers are just one piece, with knock-on needs for transportation upgrades and a modernized power grid. "We are seeing major sources of demand that are converging at the same time," Flynn said.
Justin Sinkovich at Columbia College Chicago called the all-displacement storyline "incomplete." AI cannot pour concrete, wire substations, or install cooling, water, and piping systems.
Pay, scale, and where the jobs are
The paychecks are real. Flynn said an apprentice technician can make $40,000 to $60,000, while seasoned electricians can top $100,000. ZipRecruiter's Nicole Bachaud is seeing even bigger numbers at the high end: the average minimum pay posted for data center jobs surged 125.1% from a year ago to nearly $208,000. "This is suggesting that highly specialized, top-tier engineering roles are pulling the overall average up drastically," she said.
The lift is not just for engineers. Bachaud said postings for welders and pipefitters are up 164% year over year. She cautioned that as more centers are built, a small sample can magnify moves, but even a few roles per site add up as construction spreads.
The growth is not even across the map, either. Houston and Birmingham stand out, and places with room to build and fewer regulatory hurdles are positioned to keep attracting projects. Translation: tradespeople in Birmingham likely face a stronger market than those in Boston.
Big checks are landing in specific states. The work has to be local. "These projects create demand for on-site workers and therefore cannot be offshored or conducted remotely, unlike manufacturing and software," he added. He also sees a longer arc where AI helps enable smarter, more distributed energy systems, spurring demand for controls, batteries, microgrids, and related infrastructure.
Cushman & Wakefield puts some bounds on the spillovers. By its estimates, each additional 100 megawatts of data center capacity supports close to 1,300 jobs in the surrounding economy, roughly $110 million in annual pay, about $344 million in total output, and around $187 million in regional output after accounting for local value added.
Local tensions, tax fights, and whether the benefits stick
The national mood is chilly. Gallup reports that 70% of Americans do not want a data center in their area, and a New York Times/Siena survey done this month found roughly two-thirds opposed across party lines. Even in a fast-building state like Texas, 57% of voters said they would oppose one in their community, per the Texas Politics Project, a shift some analysts say could aid Democrats in the state. According to Data Center Watch, resistance spreading this year has led to at least 75 projects, totaling roughly $130 billion, being stalled or pushed back.
Policy pushback is following. Texas Governor Greg Abbott recently paused grid approvals for data centers and halted new environmental permits. In New York, Governor Kathy Hochul put in place the country's inaugural state-level prohibition on adding new data centers.
When industries reshape how things are built, stay informed with the free Market Briefs that summarizes markets in five minutes
The tensions helped prompt Oracle to submit a "force majeure" filing connected to its New Mexico data center associated with the broader Stargate AI initiative, seeking to delay payment if the site is not up and running by 2028. The project has faced community pushback in the run-up to the midterms as well as environmental objections, though Oracle says it is still on schedule.
Economists warn against counting chickens too early. Michael Hicks of Shenandoah University said backlash often stems from overpromising. "Most of the backlash is the result of developers being untruthful with local officials about potential effects, developers have woefully overpromised jobs and given tax breaks that are beyond obscene," he said. Longer term, he added, "Permanent labor market effects are muted.
Brookings research comes to a similar conclusion on nuance: construction brings local gains, but the size varies widely and depends on facility type, such as hyperscalers versus colocation sites. The researchers also flagged that the policy debate is moving faster than the hard evidence.
Taxes may be the quiet headline. Hicks said that if communities tax data centers at household rates, the property base can soar. He cited Loudoun County, Virginia, where average property tax bills have fallen by $3,400 since 2020 amid a massive buildout.
That influx can lower tax bills or pay for measures to mitigate noise and stress on the electrical grid. Meanwhile, some states are offering large tax incentives that critics say are unnecessary and may boomerang.
The investor angle
In plain English: the AI buildout is a jobs engine for skilled trades and select engineers, and suppliers tied to cooling, power, and construction see real demand. But the gains are uneven by geography and face polling headwinds, permitting pauses, and election-year politics that can slow starts or stall sites.
The bottom line: watch where projects concentrate - places like Houston and Birmingham today - and track policy moves like Texas's moratorium and New York's ban. For companies exposed to data center construction, grid upgrades, HVAC, and piping, revenue timing will follow those shovels, even as long-run local job effects look strongest during the build phase.
Because local hiring and infrastructure affect many lives, get the free Market Briefs newsletter for a quick daily economic update
