The Great Infrastructure Backlash: Why AI Data Centers Have Become America’s Most Unpopular Tech Frontier

By Richard Lawler
Senior Editor, Technology, Culture, and Policy

Main Facts

A comprehensive new poll released by The New York Times and Siena University has quantified what local governments, regional utility boards, and federal lawmakers have increasingly observed on the ground: artificial intelligence and the massive data centers required to build and sustain it have become profoundly unpopular among the American electorate.

According to the survey data, which polled 1,503 likely voters in early September ahead of high-profile industry discussions regarding AI safety slowdowns, public sentiment toward the physical footprint of the tech boom is overwhelmingly skeptical. When asked whether they support or oppose the construction of massive data centers designed to power advanced artificial intelligence technology, 61 percent of respondents stated they were opposed.

By contrast, only 14 percent of participants expressed strong support for these projects, highlighting a severe enthusiasm gap for Big Tech’s sprawling physical expansion.

The backlash reveals a deep bipartisan and demographic divide, though opposition cuts across nearly every segment of society. Among voters who supported Donald Trump in the 2024 presidential election, opinion was relatively balanced, split nearly down the middle with 49 percent expressing support and 45 percent voicing opposition. However, the sentiment swung dramatically among other voting blocs: individuals who voted for Kamala Harris leaned heavily against data center construction at 74 percent, while non-voters registered 67 percent in opposition.

When researchers drilled down into the nature of that opposition, environmental and resource concerns topped the list. Thirty-two percent of opposed voters cited environmental degradation and massive water usage—essential for cooling high-density server farms—as their primary grievance. Another 21 percent pointed to the disruptive impact these massive industrial facilities have on local communities. General distrust or dislike of artificial intelligence accounted for 18 percent of opposition, while regulatory concerns and economic impacts registered at 10 percent and 9 percent, respectively.

Despite this visceral, widespread resistance, the data reveals a fascinating paradox: while voters intensely dislike the physical reality of AI infrastructure in their backyards, it remains a low-tier issue when it comes to deciding their votes in upcoming elections. When asked to rank the most important issues driving their midterm votes, AI and data centers registered at less than one percent across almost all demographics. The only notable exception appeared among younger voters aged 18 to 29, where three percent identified the issue as a primary concern—still dwarfed by traditional heavy-hitters like the broader economy, foreign policy, inflation, and partisan control of government.

Chronology of an Escalating Crisis

The growing friction between local communities and the technology sector did not materialize overnight. It is the culmination of a rapid, post-pandemic gold rush that caught municipal planners and utility grids completely flat-footed.

• Late 2022 – 2023: The Generative AI Boom Ignites: Following the public rollout of OpenAI’s ChatGPT, tech giants including Microsoft, Google, Amazon, and Meta launched an unprecedented race to secure computing power. Announcements for multibillion-dollar server complexes flooded suburban and rural markets across Northern Virginia, Texas, Ohio, and the Pacific Northwest.
• March 2026: Early Warning Signs in National Polling: An NBC News poll published in March 2026 delivered an alarming indicator for the tech industry, revealing that public sentiment toward AI infrastructure and its corporate architects was polling lower than Immigration and Customs Enforcement (ICE)—a historically polarizing political lightning rod.
• May 2026: Gallup Confirms the NIMBY Wave: A comprehensive Gallup poll released in May 2026 confirmed that opposition to local data center construction had reached a critical mass, finding that roughly seven in ten Americans actively opposed the building of AI data centers in their immediate areas.
• September 2026: Midterm Politics Converge: Ahead of the midterm elections, the New York Times/Siena poll captured the political fallout. Politicians who once eagerly welcomed tech campuses with tax breaks and ribbon-cuttings suddenly found themselves pivoting to oppose them, responding directly to furious local constituencies. Concurrently, discussions around a "Big Tech AI slowdown" or safety pact began circulating among industry leaders, though these moves were driven more by internal capital expenditures and energy constraints than public relations.

Supporting Data and Demographic Breakdown

The polling numbers paint a granular picture of how different parts of the American population view the physical manifestation of the artificial intelligence revolution.

The Core Grievances of Opponents

For the 61 percent of voters who oppose data center construction, the motivations are deeply rooted in tangible, local impacts rather than abstract philosophical debates about artificial intelligence:

• Environmental and Water Usage (32%): Modern AI training clusters require millions of gallons of water daily for cooling towers, placing immense strain on local municipal water systems and aquifers. Furthermore, the massive electricity draw—often requiring the equivalent output of entire nuclear or coal-fired power plants—threatens to reverse decarbonization goals.
• Local Community Impact (21%): Residents frequently cite heavy construction noise, massive industrial architecture altering rural or suburban skylines, light pollution, and a dramatic increase in heavy truck traffic.
• Distrust of AI (18%): A foundational skepticism toward the tech industry’s promises, privacy violations, and job displacement feeds directly into opposition against the physical infrastructure that enables it.
• Regulatory & Economic Concerns (19% combined): Concerns that local taxpayers will foot the bill for electrical grid upgrades, road repairs, and water infrastructure while tech companies enjoy multi-year tax abatements.

Mitigation Preferences

Among those opposed to data center expansion, the electorate is split on how government should intervene:

• 56 percent favor stringent local and federal limits, such as zoning restrictions, mandatory water-recycling requirements, and caps on energy consumption.
• 38 percent prefer an outright, total ban on new data center construction until grid and environmental safeguards can be guaranteed.

Partisan Trust and Midterm Salience

When asked which political party they trust to handle the complex issues surrounding artificial intelligence and data infrastructure, respondents showed traditional partisan alignment rather than a unified consensus. Overall, among a sample subset of 554 respondents who weighed in on party competency:

• 42 percent favored Republicans.
• 40 percent favored Democrats.
• 18 percent remained undecided or refused to answer.

Despite the intensity of local opposition, the issue has struggled to break through as a primary ballot-box driver. While younger voters (ages 18–29) show a slightly higher sensitivity to the topic—with 3 percent ranking it among their top issues—the vast majority of voters remain hyper-focused on macroeconomic realities: inflation, job security, healthcare costs, and traditional partisan battles.

Official Responses and Political Realignment

The stark realities revealed by the polling data have triggered a frantic scramble among elected officials at every level of government. Politicians who spent the early years of the AI boom offering generous tax exemptions and streamlined permitting processes are now reading the room and executing sharp policy reversals.

In states like Virginia—home to "Data Center Alley" in Loudoun County—lawmakers are facing intense pressure from homeowner associations and environmental groups demanding moratoriums on new builds. Similar resistance is playing out in rural communities across Georgia, Ohio, and Arizona, where local zoning boards are routinely ambushed by packed town hall meetings filled with angry residents.

"Politicians are responding directly to what they hear at the fence line," says a policy analyst specializing in municipal tech infrastructure. "When 61 percent of your constituents tell you they oppose these facilities, and nearly 40 percent want an outright ban, the political calculus changes overnight. What used to be viewed as a clean, high-tech economic engine is now perceived as an industrial polluter that consumes infinite electricity and water while employing very few local residents once construction wraps up."

Federal lawmakers are also beginning to take notice. While Congress remains deeply fractured on comprehensive AI safety and regulation, bipartisan coalitions are forming around the need for federal oversight of the electrical grid. The sheer volume of energy demanded by AI data centers threatens to destabilize regional transmission organizations (RTOs), forcing utility companies to keep aging fossil-fuel plants online longer than planned—a reality that infuriates both climate-conscious progressives and energy-conscious conservatives.

Tech executives, meanwhile, have largely maintained a public relations silence regarding the grassroots backlash, choosing instead to focus on technological workarounds. Companies are increasingly exploring proprietary nuclear, geothermal, and advanced solar power purchase agreements to fuel their data centers independently, hoping to sidestep local utility battles and public outrage over grid reliability.

Implications for the Future of Tech and Society

The deep chasm between public aversion to AI infrastructure and the tech industry’s insatiable appetite for compute capacity carries profound implications for the trajectory of the artificial intelligence revolution.

1. The Infrastructure Bottleneck is Real

For all the billions of dollars venture capitalists and big tech firms are pouring into algorithmic development, the physical world is pushing back. If two-thirds of the population views data centers with suspicion or outright hostility, scaling up to Artificial General Intelligence (AGI) or handling ubiquitous enterprise AI workloads will face severe bottlenecks. Permitting battles will grow longer, more litigious, and politically toxic.

2. The Energy Crisis Will Dictate AI’s Speed

The environmental footprint of AI—specifically its insatiable thirst for water and electricity—is no longer a theoretical talking point for environmentalists; it is a hard physical limit. Tech companies will be forced to decentralize, moving facilities to remote, unpopulated regions, or invest massively in behind-the-meter green energy solutions. Without secure, clean power sources that do not burden local taxpayers, public backlash will likely force federal or state-level legislative caps.

3. A Disconnect Between Macro Politics and Local Reality

The polling data highlights a persistent disconnect: voters hate data centers in their towns, but they do not yet view AI as a defining national political issue. However, as local zoning disputes escalate into national political flashpoints, that dynamic could shift. If an AI data center expansion directly results in rolling brownouts, skyrocketing utility bills, or depleted municipal water tables in key swing districts, the issue could quickly transition from a local nuisance to a formidable wedge issue in federal elections.

As the tech industry races toward its next generation of models, it must confront a sobering reality: you cannot build the digital future if the physical communities anchoring it refuse to let you in.