The American backlash against data centers is no longer a narrow dispute over zoning. It is becoming a political coalition built around electricity prices, water, pollution, land rights and distrust of the companies driving the AI boom. Wired’s reporting from Georgia shows how activists from very different ideological camps are finding common ground around the physical infrastructure behind artificial intelligence.
The scale of the opposition is notable. A recent Gallup poll cited by Wired found that 70 percent of Americans would somewhat or strongly oppose a data center in their area. That sentiment is drawing national politicians into fights once handled mainly by planning boards and utility regulators.
AI infrastructure has become a local cost
For technology companies, hyperscale facilities are the machinery required to train models and serve millions of users. For communities, they are large industrial projects that demand power, water, transmission lines and favorable permits. The two descriptions refer to the same buildings but create very different measures of value.
Residents often encounter the costs before they see a clear local benefit. New power generation can affect air quality, transmission can require private land, and utility investments can create concern about household bills. Wired points to Elon Musk’s xAI site in Memphis, where unpermitted gas turbines were reportedly brought into a majority-Black neighborhood already burdened by pollution. That example has become part of a wider argument that AI expansion can intensify existing environmental inequality.
Developers may promise construction work, tax revenue and economic growth. Those are projections that depend on specific agreements and should be weighed against the number and durability of jobs, public subsidies and infrastructure obligations.
Opposition crosses familiar party lines
Data center resistance does not fit cleanly into the American left-right divide. Environmental justice groups focus on pollution and resource use. Property-rights advocates object to eminent domain and transmission corridors. Populist conservatives describe the buildout as an alliance between government and powerful technology companies, while labor and consumer advocates question who pays for new capacity.
Republicans Marjorie Taylor Greene and Thomas Massie were among the national politicians Wired identified as early critics of data center development. Greene argued that using eminent domain to connect facilities was incompatible with conservative principles. Elsewhere, progressive figures have challenged corporate water and energy demands. The shared target can conceal major disagreements about climate policy, regulation and public investment, but it makes the coalition politically difficult to dismiss.
The movement also benefits from the industry’s abstract language. Terms such as compute, cloud and hyperscale can make infrastructure sound weightless. Local campaigns replace that abstraction with turbines, substations, wells and monthly bills.
The AI race now needs public permission
Companies can respond by treating opposition as a communications problem, but public distrust is rooted in material questions. Communities want enforceable answers about water withdrawals, power sourcing, emissions, noise, tax concessions and who pays when demand forecasts prove wrong.
Greater transparency would not guarantee approval. It would, however, let residents compare claims across projects and hold developers to measurable commitments. Regulators also need independent capacity to test corporate projections rather than relying on the urgency of the AI race as justification.
The political shift matters because infrastructure determines the practical speed of AI expansion. Chips and models receive attention, but neither works without facilities connected to energy and networks. If developers cannot establish legitimacy where those facilities are built, national ambitions will collide with local veto points.
Data centers are not breaking politics by themselves. They are revealing a deeper conflict over how technological benefits and industrial burdens are distributed. AI companies have spent years explaining what their systems might eventually do. Their next challenge is more immediate: persuading communities that the infrastructure serving those systems deserves a place in everyday landscapes.
