A joint mapping project traces both the AI build wave and the organized local pushback, and finds that the rural counties taking the build weren't staffed to negotiate what's arriving.
The AI infrastructure boom is moving off the congested grids of Northern Virginia and into the corn belt. About 6% of U.S. data centers currently in operation sit in rural counties. Of the facilities now under construction, roughly 12% are slated for rural sites. The counties welcoming them were not built to negotiate what a warehouse-scale facility actually costs.
That gap is the heart of a new mapping project published by the Daily Yonder and GPB News, and it has two halves. The first is the geography of the current AI build wave. As demand pulls hyperscale campuses out of congested metros, rural counties with available land, cheap power, and willing local officials are absorbing a disproportionate share of new projects, even though they only hold a small slice of today's operating footprint. The second half is the local-fight map, the communities pushing back when a developer, a utility interconnection request, or a proposed tax increment shows up at a county board meeting that meets twice a month.
The data comes from FracTracker Alliance's Open U.S. Data Centers Tracker, the most complete public dataset of its kind. It is crowd-sourced and updates monthly. The rural classification uses the federal Office of Management and Budget's Metropolitan Statistical Area definition. Any county outside an MSA counts as rural. That framing matters, because some "rural" data centers sit on the edges of metro regions, and the project's authors disclose that county-level line when readers ask what "rural" means here.
What the map shows at the structural level is a negotiation gap. A large warehouse-scale facility typically comes with a multi-decade tax abatement, a request for new high-voltage transmission, a decommissioning bond sized for an industrial site, and water and noise mitigation that small planning offices are not staffed to evaluate. The metros where hyperscale grew up first spent the last decade staffing up for this kind of negotiation. Rural counties, by and large, have not.
Wisconsin's Port Washington and Ozaukee County case shows what that mismatch looks like in practice. American Transmission Co. has proposed an Ozaukee County distribution-interconnection project to serve a data center load, and the Wisconsin Public Service Commission is reviewing whether the line is needed and who pays for it. The state Department of Natural Resources has launched a Port Washington environmental impact assessment. Residents have organized around water intake, setback distance, and grid-cost allocation. None of those fights have been adjudicated yet. The docket is open and the commission has not ruled.
Nationally, public opinion on data centers has hardened in the past 18 months. A Gallup survey found a majority of Americans oppose data centers in their area, and a Reuters/Ipsos poll released June 11, 2026 found Americans wary of the AI-driven boom specifically. Neither poll isolates rural sentiment from the national sample, so the natural assumption, that rural opposition runs hotter because local negotiating capacity is thinner, remains a working hypothesis, not a measured finding. Tracking organizations like the Minnesota Center for Environmental Advocacy are starting to compile the legal record that would let analysts test that assumption against the actual docket pattern.
For now, the geography is the story. The current AI build wave is doing something the cloud build wave of the 2010s did not. It is concentrating capital in counties whose planning capacity has not scaled to match it. The next round of state commission dockets, county zoning votes, and utility integrated resource plans will decide whether rural counties develop the staff and the model agreements to absorb that capital on their own terms, or whether each fight has to be reinvented from scratch.