Seven out of ten Americans don’t want an AI data center built anywhere near them, according to Gallup. Nearly half are strongly opposed. That’s not a vocal minority on Reddit — that’s a supermajority. And the consequences are already real: Data Center Watch reports
$64 billion in U.SAI infrastructureinvestment at this scale has become a flashpoint few anticipated.
From Abstract Fear to Concrete Resistance #
Opposition has spread across 24 states, and the reasons go far beyond “not in my backyard.”
This isn’t one angry town hall. NPR and U.S. News report protests halting or slowing proposed facilities in Virginia, Pennsylvania, and North Carolina. Data Center Watch counts 142 activist groups organized across 24 states. The pattern looks less like scattered NIMBYism and more like the early stages of a national movement — think the anti-fracking coalitions of the 2010s, but with server racks instead of wellheads.
Electricity ratepayers are already part of this story. Opposition blends concerns about surging power costs, massive water consumption, underwhelming local job creation, privacy erosion, and the raw concentration of corporate power. Brookings highlights public anxiety around privacy intrusions, cybersecurity vulnerabilities, and racial and gender bias embedded in AI systems. The grievances aren’t abstract. They show up in utility rates and aquifer levels.
- 70% of Americans oppose local data center construction, according to Gallup
- $64 billion in projects blocked or delayed nationwide, per Data Center Watch
- 142 activist groups active across 24 states
- Only 26% of voters hold positive feelings about AI, per CFR polling - Just 10% say they’re more excited than concerned
The Cost of Getting This Wrong #
Overreaction carries its own risks — including ceding AI leadership to a geopolitical rival.
China stands to fill any vacuum left by U.S. policy overreach, according to the Council on Foreign Relations. That carries direct consequences for cybersecurity and military capability. CFR argues that squandering AI’s potential gains through excessive restriction is its own form of national risk — not a cautious hedge, but a different kind of loss.
AI tools are already embedded in daily work, smartphones, and automated public services. Outright rejection gets harder when the technology is already woven into the fabric of ordinary life. The friction is international, too — Daily Friend reports Samsung workers in South Korea threatened to strike over chip-related profits, adding a global labor dimension to the dispute.
The policy path that actually works isn’t a broad ban. It’s benefit-sharing, targeted regulation, and deploying AI to improve the public services people already depend on. Because right now, communities are absorbing the costs while someone else cashes the check.