There is a peculiar irony at the heart of America's digital infrastructure boom. The same rural communities that have waited decades for reliable internet access are now being courted by data center developers who need their land, their water, and their electrical grid. Billions of dollars in computing infrastructure are being planted in counties where residents still struggle to load a webpage.
The opportunity here is real. But so is the risk of a repeat performance—another extractive industry that takes what it needs from small towns and leaves nothing behind.
The Broadband Desert That Data Centers Could Water
Despite years of federal promises and hundreds of billions in pledged investment, a staggering number of rural Americans still lack adequate broadband. Estimates vary, but somewhere between 22 and 50 percent of rural households cannot access internet speeds that most urban residents take for granted. In parts of the rural South—Georgia, Alabama, Mississippi, Louisiana—the gap is even wider. These are communities where a student cannot reliably attend a virtual class, where a farmer cannot access precision agriculture tools, and where a small business owner cannot process a credit card transaction without driving to town.
The federal government's flagship effort to close this gap is the Broadband Equity, Access, and Deployment (BEAD) program, created under the 2021 Infrastructure Investment and Jobs Act. The program allocated $42.5 billion to expand high-speed internet to unserved and underserved areas, with every state and territory receiving at least $100 million. Georgia received $1.3 billion. Alabama received $1.4 billion. Louisiana received $1.36 billion.
By early 2026, states have submitted their final proposals and construction is beginning. But the BEAD program has not been without controversy. Rule changes in mid-2025 introduced a "technology-neutral" approach that expanded the role of satellite and fixed wireless alongside fiber, and roughly $21 billion of the original allocation remains unspent. Whether the money reaches the communities that need it most—and whether it arrives as fiber or something lesser—remains an open question.
Enter the data center.
Why Data Centers Need What Rural America Has
The explosion of artificial intelligence has created an unprecedented demand for computing power. U.S. data center capital expenditures exceeded $400 billion in 2025. Hyperscale operators—Amazon, Google, Microsoft, Meta—are racing to build campuses that can house hundreds of thousands of servers, and they need three things that rural America has in abundance: land, power, and relative quiet from regulatory opposition (or so they hope).
The average data center site in 2024 covered roughly 224 acres—a 144 percent increase from 2022. Some hyperscale campuses exceed 1,000 acres. That kind of footprint is simply unavailable in the urban corridors where most data centers have historically been built. Northern Virginia's Loudoun County, home to the densest concentration of data centers on Earth, is running out of room. So developers are looking outward: to the Piedmont region of Virginia, to central Georgia, to the rural Midwest, to anywhere the land is flat, the power lines are accessible, and the county commissioners are willing to talk.
This is where the broadband connection becomes critical. Every data center requires high-capacity fiber-optic routes to function. When a hyperscale facility moves into a rural county, it does not arrive alone. It brings fiber—miles and miles of it—running from the facility to the nearest internet exchange points and network hubs. That fiber is the backbone of broadband infrastructure. And in communities that have never had it, the arrival of a data center could, in theory, provide the "middle mile" connectivity that makes last-mile broadband economically viable for the first time.
The Loudoun County Lesson
Loudoun County, Virginia, offers the most instructive case study in the country. The county's data center industry now generates more than $1 billion in annual local tax revenue—funding roughly one-third of the county budget. Since 2012, residential property tax rates have dropped by 48 cents per $100 of assessed value, saving the average homeowner thousands of dollars annually. The data center industry, in effect, subsidizes the community.
But Loudoun County is also investing heavily in rural broadband. A partnership with All Points Broadband is extending fiber-to-the-home service to approximately 8,600 previously unserved rural households and businesses, funded in part by a Virginia Telecommunication Initiative (VATI) grant. By early 2026, Dominion Energy had completed 168 miles of middle-mile fiber construction in the county, with over 105 miles of last-mile fiber connecting directly to homes.
The lesson is not that data centers automatically produce broadband. They do not. The lesson is that when a community negotiates deliberately—when it uses the leverage of its land and its resources to demand infrastructure investment in return—data center development can become a catalyst for connectivity. Loudoun County did not get lucky. It got strategic.
The Edge Data Center Opportunity
Not every rural community is going to land a hyperscale campus. But the rise of edge computing—smaller data centers positioned closer to where data is generated and consumed—offers a different and perhaps more relevant opportunity for small towns.
Edge data centers reduce latency, which is the delay between a user's request and the server's response. For rural communities, lower latency means better telemedicine, more reliable remote work, smoother online education, and the ability to support precision agriculture and IoT-connected farming equipment. These are not luxuries. In a county where the nearest hospital is 45 minutes away, a telemedicine connection that actually works is a matter of life and death.
Edge facilities are smaller—often fitting inside a single building or even a shipping container—and their power and water demands are a fraction of what a hyperscale campus requires. They can be deployed in existing commercial spaces, on municipal property, or co-located with other infrastructure like water treatment plants or electrical substations. For a town of 5,000 people, a well-placed edge data center could mean the difference between digital irrelevance and economic viability.
Rural broadband operators and electric cooperatives are already exploring this model. CoBank, the cooperative bank that serves rural America's infrastructure lenders, has published extensive research on the potential for co-ops to become data center hosts—turning their existing fiber and electrical assets into platforms for edge computing. It is a model that keeps ownership and economic benefit local, rather than ceding both to a hyperscaler based in Seattle or Menlo Park.
What Can Go Wrong—and Usually Does
We would be irresponsible not to name the risks. The pattern of data center development in rural America has, to date, been characterized more by extraction than by partnership.
Developers frequently choose rural locations for their low costs, not for any commitment to local economic development. Tax incentive packages—often negotiated behind closed doors with confidentiality agreements that prevent public scrutiny—can hollow out the promised fiscal benefits. A community that offers a 20-year tax abatement in exchange for a data center may find that it has traded its most valuable resource (land and water access) for a handful of permanent jobs and a facility that could be mothballed when the technology cycle shifts.
Permanent employment at data centers is routinely overstated. A facility that employs 2,000 construction workers for 18 months may require only 30 to 50 full-time operational staff. Many of those positions require specialized skills that local workers may not possess, leading developers to import talent rather than hire locally. Without enforceable workforce development commitments—partnerships with community colleges, apprenticeship programs, technical training pipelines—the jobs promise evaporates.
And then there is the resource question. A single modern data center can consume as much electricity as 100,000 homes. Larger AI-focused facilities can use 20 times that. In rural counties with limited grid capacity, a data center can drive up utility rates for every other customer on the system. Water consumption follows the same pattern: millions of gallons per day, drawn from the same aquifers that feed municipal water systems, irrigate farms, and sustain ecosystems.
The September 2025 introduction of the federal "Unleashing Low-Cost Rural AI Act" acknowledged these risks, mandating that the Departments of Agriculture, Interior, and Energy study the impact of data center expansion on rural energy supply, reliability, and consumer costs. It is a start. But a study is not a safeguard.
What We Believe Small Towns Should Demand
At the EPR Foundation, we believe digital infrastructure is essential to the future of rural America. We also believe that small towns are capable of negotiating from a position of strength—if they know what to ask for.
Broadband as a condition of approval. Any data center permitted in a rural county should be contractually required to extend broadband access to surrounding communities. The fiber is coming anyway. The only question is whether it stops at the data center's property line or continues to the homes and businesses nearby. Community benefit agreements should include specific broadband commitments with enforceable timelines and speed benchmarks.
Transparent resource accounting. Before a permit is granted, communities should have access to detailed projections of electricity consumption, water withdrawal, and grid impact—reviewed by independent engineers, not just the developer's consultants. New Jersey's requirement that data centers report electricity and water usage quarterly should be the national standard.
Local workforce pipelines. Tax incentives should be contingent on verifiable local hiring commitments and funded partnerships with community colleges and technical schools for training in electrical systems, HVAC, networking, and facility management.
Revenue that reflects the burden. Property tax revenue from a data center should be proportional to the infrastructure demands it imposes. If a facility requires a new substation, upgraded water lines, or road improvements, those costs should be borne by the developer—not by the county's existing taxpayers.
The fiber that connects a data center to the outside world could just as easily connect a rural school, a family farm, and a small-town hospital. The question is whether anyone bothers to make the connection.
Rural America has been promised connectivity before. The telephone took decades to reach every farmhouse. Rural electrification required an act of Congress. Broadband has followed the same agonizing trajectory—always coming, never quite arriving. The data center boom represents a genuine, if imperfect, opportunity to break that cycle. But only if small towns refuse to be passive hosts and instead insist on being partners.
The digital future does not have to bypass the places that built this country. But it will—unless those places demand a seat at the table before the concrete is poured.