For nearly 40 years, John Steinbach has called Manassas, Virginia, home. But in January 2026, opening his monthly electricity bill brought an unwelcome shock: a charge of $281, representing a dramatic spike from the roughly $100 he had paid just the month before. "It’s just so far beyond any bill that I’ve ever had," Steinbach told Consumer Reports, capturing the growing anxiety shared by millions of Americans. What worries residents most is not merely a single high month, but the trajectory of future rates as a rapidly expanding fleet of artificial intelligence data centers plugs into the local electrical grid.
Steinbach is far from alone in his apprehension. A Gallup poll conducted in March 2026 revealed that seven in 10 Americans oppose the construction of AI data centers in their immediate geographic areas, with nearly half—48%—voicing strong opposition. Notably, this skepticism transcends traditional political divides, drawing majority opposition from Democrats, Republicans, and independents alike.
This widespread grassroots resistance has translated into significant administrative hurdles for the tech industry. According to data from Data Center Watch, local pushback disrupted approximately 120 projects during the first half of 2026 alone. At least 45 of those developments, valued at an estimated $68 billion, stalled out during the second quarter. Across the United States, roughly 379 jurisdictions have enacted moratoriums or outright bans on data center construction.
Municipalities are moving swiftly to put the brakes on development. In August 2026, the Indianapolis City Council voted 23–1 to bar new data center construction through the end of 2027. Similarly, Charlotte officials approved a 150-day moratorium following a municipal survey that revealed 78% local opposition. According to the Brookings Institution, political candidates across the ideological spectrum are now actively running campaign advertisements opposing these industrial facilities.
Public debates are typically framed around whether data centers should be built at all. However, a much more complex and contentious question underlies these local battles: who pays, and who decides? The costs of these massive facilities tend to land locally and immediately, while the economic benefits often materialize regionally or nationally at a much later date. Crucially, the regulatory rules governing how these costs and benefits are split were written long before AI-scale electricity demand ever existed.
Why AI Changed the Math
Traditional cloud computing data centers were built around general-purpose servers whose power loads fluctuate naturally alongside user traffic. In contrast, modern AI clusters pack power-hungry graphics processing units much more densely, generate significantly higher heat loads per rack, and draw immense amounts of power around the clock during intensive training runs. This profound technological shift dramatically alters what a single campus demands from local electrical grids and municipal water systems.
According to a report prepared for the U.S. Department of Energy by Lawrence Berkeley National Laboratory, U.S. data centers consumed approximately 176 terawatt-hours of electricity in 2023, accounting for roughly 4.4% of the nation’s total power generation. The central point of contention, however, centers on future growth.
A power report published by Bloom Energy projects that U.S. data center IT loads will nearly double, surging from approximately 80 gigawatts in 2025 to about 150 gigawatts by 2028. While this projection comes from an energy company that manufactures on-site power systems—giving it a clear commercial stake in the outcome—it highlights a very real infrastructure bottleneck. Traditional utilities currently expect to deliver required grid capacity 1.5 to 2 years later than developers intend to deploy their projects. To bridge this gap, developers project that roughly a third of all data centers will run entirely on on-site power by 2030, effectively building their own private power plants when the public grid fails to keep pace.
Electricity Bills: Who Pays for the Next Grid
The financial anxieties surrounding power consumption are vividly illustrated by market dynamics within PJM Interconnection, the regional grid operator serving 13 states and Washington, D.C. PJM runs capacity auctions designed to pay power plants to remain available during periods of peak electricity demand.
The clearing price in these auctions skyrocketed from $28.92 per megawatt-day for the 2024–2025 period to $269.92 for 2025–2026, eventually hitting regulatory price caps in subsequent auctions at $329.17 and $333.44. PJM’s independent market monitor attributed 63% of the 2025–2026 price increase—amounting to roughly $9.3 billion—directly to rising data center load.
Other economic indicators point to similar pressures. Energy Information Administration data shows that residential electricity prices in Virginia rose by about 13% over a 12-month period. A January 2026 survey conducted by Global Strategy Group and the Chesapeake Climate Action Network Action Fund found that nearly three-quarters of Virginia voters blamed data centers for these hikes. Looking further ahead, a peer-reviewed modeling study conducted by researchers from NC State, Carnegie Mellon, and two other universities projects that wholesale electricity prices could rise between 6% and 29% nationally by 2030 compared to a baseline without data center expansion, with increases reaching up to 57% in the hardest-hit regions.
Conversely, the technology industry argues that massive corporate customers typically help lower overall rates. An Electric Power Research Institute study examining the 2015-to-2024 period concluded that each doubling of data center capacity correlated with a 3.5% reduction in average retail prices, as large and steady power consumption spreads a utility’s fixed operational costs across a broader customer base.
An Amazon-funded study by consulting firm E3 similarly found that Amazon’s utility payments in four distinct territories met or exceeded the actual costs required to service its data centers, leaving an estimated $3.4 million in net surplus per site. A separate E3 report funded by the Data Center Coalition estimated that load growth accounted for roughly 50% of the 2025–2026 PJM price spike—notably lower than the independent market monitor’s figure. Furthermore, the free-market Institute for Energy Research published findings indicating no statistically significant correlation between a state’s data center count and its retail electricity prices.
Empirical evaluations often land on a cautious "not yet." The Rutgers New Jersey State Policy Lab found no statistically significant effect on utility-level residential bills through 2024, though researchers added the critical caveat that "’not yet’ is not the same as ‘nothing to worry about.’" Historical economic benefits have largely relied on having spare electrical grid capacity. If infrastructure is built for anticipated demand that ultimately fails to materialize, the massive fixed costs of expanding grid capacity will inevitably be absorbed by a smaller population of everyday ratepayers.
Water: National Totals, Local Stress
Beyond electricity, water consumption represents another major flashpoint. Water-cooled data centers can consume millions of gallons on a single hot summer day, putting immense strain on municipal resources, particularly in arid regions. Arizona, for instance, faces a federally mandated plan that will cut its allocation from the Colorado River by 760,000 acre-feet—a reduction of roughly 27% to 30%.
This environmental context heavily influenced local government decisions. In August 2025, the Tucson City Council unanimously rejected Amazon’s proposed Project Blue amid intense community pressure, while Chandler officials similarly voted down an AI data center proposal 7–0 that December. In Florida, the vice mayor of Orange County proposed a one-year moratorium on new facilities, pointing to a projected regional groundwater shortfall of 96 million gallons per day by 2045.
Viral concerns regarding water usage—such as the widely cited estimate that each 100-word AI prompt consumes about half a liter of water, derived from a GPT-4-era calculation by UC Riverside researchers—often clash with corporate metrics. Google’s internal figures for a median Gemini text prompt indicate on-site consumption of about 0.26 milliliters, or roughly five drops, though that figure excludes off-site thermoelectric water consumption.
At a national level, Lawrence Berkeley National Laboratory estimates that U.S. data centers directly consumed about 17.5 billion gallons of water in 2023, a figure projected to double or quadruple by 2028. Analysts note that this accounts for a small fraction of overall freshwater usage; figures cited by The Dispatch place data centers at under 0.5% of total U.S. freshwater consumption, comparing the water footprint of facilities like Google’s Iowa campus to standard golf courses. Furthermore, an Arizona State University report found no measurable rise in central Arizona power suppliers’ water usage despite regional data center expansion, noting that agriculture continues to account for 72% of the state’s total water consumption.
However, national averages frequently obscure severe local stress. In Botetourt County, Virginia, a planned Google campus could consume between 2 million and 8 million gallons of water daily in a rural county where many residents rely on residential wells. To accommodate the facility, the county is allocating $300 million toward a new regional water supply, funded in part by the project’s anticipated tax revenue.
Jobs and Tax Revenue: A Strong Case and a Thin One
The tech industry’s most compelling argument centers on municipal tax revenue. Data centers across Northern Virginia generated an estimated $1.3 billion in local property taxes in 2024, with total supported local revenue exceeding $2 billion, according to data from the Chamber of Progress.
Quincy, Washington, is frequently cited as a primary success story. In that small agricultural town, roughly 30 data centers account for an estimated 57% of total property tax revenues. This expanded tax base has helped fund a $120 million high school, a local hospital, new police and fire stations, and a $30 million water reuse facility built in partnership with Microsoft. In Texas, industry figures reported by the Texas Tribune claimed 61,060 direct jobs and $3.5 billion in state and local tax contributions in 2023, with proponents arguing that a proposed $10 billion campus would inject $50 million annually into Lacy Lakeview, a town of roughly 7,000 residents. Even legislative auditors in Virginia ranked the state’s data center tax exemption as its second-most productive economic incentive, generating $6.1 million in personal income for every $1 million in forgone tax revenue.
Conversely, the direct employment generated by these facilities is comparatively modest. A Brookings Institution analysis of approximately 1,500 data centers found that a county’s first major facility typically adds between 100 and 200 jobs over a decade, with no discernible impact on local wage levels. State incentive programs often reflect these modest job figures; Texas grants sales tax exemptions to facilities spanning at least 100,000 square feet that create just 20 jobs, while Maryland requires five jobs and a $2 million investment in targeted areas. While construction jobs provide a temporary economic boost, employment levels drop off sharply once building is complete.
Recognizing this imbalance, state governments are adjusting their policies. Lawmakers in Illinois, Massachusetts, Nebraska, Nevada, and Ohio paused or reworked data center tax breaks, while Arizona temporarily froze new awards for three years. Virginia preserved its sales tax exemption but instituted a temporary electricity tax of 1.1 cents per kilowatt-hour on data centers to run from July 2026 to June 2028, capped at $600 million annually.
Air Pollution and On-Site Power: The Memphis Case
Memphis serves as an extreme example of what can happen when tech companies bypass conventional utility grid timelines. To bring its Colossus 1 supercomputer online in 2024, xAI deployed trailer-mounted methane gas turbines in South Memphis—reaching as many as 35 units, according to aerial imaging from the Southern Environmental Law Center—near Boxtown, a community already burdened by historic industrial pollution. For its Colossus 2 project, the company installed additional turbines just across the state line in Southaven, Mississippi, totaling 46 units by May 2026.
Initially, classifying the power units as non-road engines allowed xAI to avoid standard air permitting requirements until an EPA rule issued in January 2026 clarified that large stationary gas turbines require formal environmental permits. SpaceX, which acquired xAI in February, ultimately took responsibility for 69 turbines, agreed to phase out the unpermitted units by July 2027, and began construction on a permanent 1.2-gigawatt natural gas power plant. Civil rights organizations, including the NAACP represented by the Southern Environmental Law Center and Earthjustice, subsequently filed a federal lawsuit under the Clean Air Act.
The project has drawn federal backing as well. In June, the Department of Justice intervened on xAI’s side, arguing that the unpermitted turbines constituted a matter of national economic and energy security. While third-party air monitoring tests indicated that ten measured pollutants remained below dangerous thresholds, environmental groups noted that critical compounds like ozone were not tested and that prevailing winds blew away from monitoring equipment during testing days.
Independent scientific analyses of the Memphis site have produced mixed conclusions. A University of Tennessee, Knoxville analysis of satellite data for TIME magazine found that average nitrogen dioxide levels rose 3% following June 2024, with peak concentrations surging up to 79% directly adjacent to the site and 9% in Boxtown. Meanwhile, University of Memphis researchers modeling local air quality calculated a roughly 1% rise in fine particulate matter alongside minimal shifts in broader satellite data, though they acknowledged their methodology could not detect several hazardous compounds. While the average environmental impact may appear limited, critics emphasize that these emissions are added to an airshed that already exceeds federal air quality standards.
Local Control and the Demand Risk
Much of the broader public backlash stems from opaque municipal processes. Before Applied Digital’s $3.6 billion Louisiana campus was publicly announced as Delta Forge 1, negotiations occurred under non-disclosure agreements signed by state senators, local assessors, and project approval board members. Conversely, other Louisiana projects proceeded transparently when local leaders refused secrecy agreements. Lawmakers have since introduced legislation in Michigan, Oklahoma, Kentucky, and Ohio to ban non-disclosure agreements for such economic development deals, while Pennsylvania Governor Josh Shapiro barred state agencies from signing them.
Zoning regulations can similarly sideline local residents from public input processes. In Aurora, Colorado, Vantage’s proposed 500-megawatt Trident campus was slated for administrative approval without public hearings because existing zoning already permitted data centers. The campus, positioned opposite a residential development, is designed to draw roughly 11 times the electrical power consumed by Denver International Airport.
State governments are increasingly stepping in to oversee local disputes. In August, Texas Governor Greg Abbott ordered a comprehensive state audit of every data center seeking a grid connection, following up in September by halting state-level permits for data center projects until the audits are complete. As of mid-2026, lawmakers in at least 14 states had introduced moratorium bills to pause data center development while long-term regulatory frameworks are evaluated.
Both critics and industry boosters share a foundational assumption that the projected demand for AI infrastructure will materialize exactly as forecast. McKinsey’s widely cited estimates project global data center capital spending to reach approximately $6.7 trillion by 2030, and both proponents and opponents frequently treat this trajectory as inevitable.
If that demand falls short, however, the financial fallout will not simply vanish. Utilities that invest billions in generation and transmission infrastructure for projects that are later downsized or canceled must still recover those capital costs from remaining ratepayers, typically by spreading fixed expenses across fewer customers. This risk has driven state regulators to institute stricter financial protections; Virginia regulators now require the largest corporate customers to pay for at least 85% of their contracted transmission and distribution demand, and 60% of generation demand, whether they utilize the power or not.
As statehouses across the country weigh new regulatory frameworks, the central dilemma remains unresolved. The immediate costs of data center expansion—spanning capacity charges, local water stress, air permitting battles, and land use disputes—fall squarely on local communities in the near term. Conversely, the economic benefits tend to accrue regionally or nationally over much longer periods. Finding a sustainable path forward will depend on whether policymakers can successfully modernize utility and zoning rules to ensure that the next wave of infrastructure development comes with a fair and transparent distribution of costs.
