AI datacenters are becoming a political and economic flashpoint, and opposition is growing on both sides of the political aisle. Developers plan to invest $8tn to build them in the coming years, a sum roughly equal to a quarter of our country’s entire GDP.
That enormous investment promises more computing power for artificial intelligence, but it also raises a direct question about who benefits. AI CEOs believe the technology will let businesses cut labor costs and increase profits, while workers and communities face concerns about jobs, public resources and the long-term value of datacenter expansion.
A Massive Investment With Few Long-Term Jobs
Datacenters require major construction projects, but they produce a minuscule number of long-term jobs. That gap sits at the heart of the growing debate: communities can host expensive facilities and provide public support, yet receive little permanent employment in return.
A datacenter in Rockland county shows how stark that imbalance can become. The facility received $77m in tax breaks and created only one permanent job. The numbers have fueled concerns that public money can support infrastructure without delivering a matching economic benefit for local workers.
Virginia hosts more datacenters than any other state, giving it a clear position at the center of this debate. A study found that datacenters in Virginia produce less than 1% of the permanent jobs per dollar invested, a figure that challenges the idea that these projects create broad, lasting employment.
The economic question reaches beyond the facilities themselves. If businesses use artificial intelligence to cut labor costs and increase profits, the datacenters powering that technology may help build a system that needs fewer workers even as it attracts vast sums of investment.
Unions Face a Choice Over AI Expansion
Labor unions now face a conflict between construction work created by datacenter expansion and the jobs that artificial intelligence may eliminate. The North American Building Trades Unions, also known as NABTU, signed agreements with Blackrock, Meta and OpenAI to collaborate on data center expansion.
Those agreements connect union labor to the physical growth of AI infrastructure. Union members will build factories of computerized labor that could put millions of workers out of jobs, creating a tension between employment today and the future of work that these facilities may help create.
The challenge is not limited to one trade or one industry. AI CEOs believe businesses can use the technology to reduce labor costs, so the same expansion that creates construction opportunities can also support a wider effort to replace human labor. That makes coordination across labor unions central to the debate over datacenters.
One warning captures the concern in a single sentence: “Our data fuels their algorithms, our taxes fund their development and our communities host their hardware.” The statement links the technology’s inputs, public support and physical footprint, placing workers and communities at the center of the conversation.
Calls for a Moratorium and New Rules
Political pressure has already produced calls for a pause. The Artificial Intelligence Data Center Moratorium Act was introduced in Congress by Bernie Sanders and Alexandria Ocasio-Cortez, putting datacenter growth inside a national debate over regulation and worker protection.
New York governor Kathy Hochul signed a statewide moratorium on data center permits, showing that the dispute is moving from public criticism into government action. A moratorium creates space to examine the economic impact of these projects before more facilities receive approval.
The facts driving that push are direct: developers plan to spend $8tn, one Rockland county datacenter received $77m in tax breaks and created one permanent job, and Virginia datacenters produce less than 1% of permanent jobs per dollar invested. Together, those figures raise a basic test for future projects: do the jobs and public benefits match the scale of the investment?
That test also reaches union strategy. NABTU’s agreements with Blackrock, Meta and OpenAI place building trades unions inside the expansion effort, while concerns about millions of potential job losses point toward a wider labor response. The debate is no longer only about whether datacenters will be built, but who will control their growth and who will carry its costs.
AI datacenters promise more computing capacity, yet their economic record has become impossible to separate from questions about jobs, taxes and community power. With a congressional moratorium proposal, a statewide permitting moratorium in New York and growing opposition across the political aisle, the next phase of AI expansion will face demands for rules that protect workers before construction moves ahead.
