AI datacenter opposition groups surged to 430 from 76 in 2025, signaling community acceptance is emerging as site-selection risk.
Organized opposition to US data center development is growing rapidly as the industry embarks on one of the largest AI infrastructure buildouts in its history, raising questions about whether community acceptance is becoming a critical factor in where—or if—projects get built.
DataCenterOpposition.com reports 430 local groups opposing data center projects across more than 40 states, up from 268 groups in April and roughly 76 at the end of 2025. These groups collectively represent more than 525,000 members—nearly seven times the level tracked six months ago.
This growth coincides with hyperscale developers announcing AI campuses measured in gigawatts, utilities revising long-term demand forecasts upward, and transmission planners pursuing multibillion-dollar infrastructure investments to accommodate projected AI electricity demand. Developers are expanding beyond traditional markets in search of available power, land, and utility capacity. Bloomberg reported that more than half of US data center projects planned for 2026 are expected to face delays from permitting, zoning, power procurement, and equipment shortages. Data Center Watch reported that opponents blocked or delayed at least 75 projects representing roughly $130 billion during the first quarter of 2026—the highest quarterly total since tracking began in 2023.
A June survey of 156 hyperscalers, colocation providers, neoclouds, developers, and chip companies conducted for Bloom Energy found that while power remains the industry's top site-selection factor, "community scrutiny" has emerged as one of the fastest-growing barriers to new development. Respondents identified higher local electricity prices, water consumption, and grid reliability as community concerns most likely to influence projects.
Recent reporting suggests local resistance is becoming an increasingly consequential factor in AI infrastructure deployment. Within a week, two high-profile Virginia projects unraveled: QTS ended its pursuit of the Prince William Digital Gateway campus after years of litigation following a successful legal challenge to the project's rezoning. Days later, the Dulles South Innovation Center proposal in Prince William County—envisioning as much as 43 million square feet of data center development—collapsed after years of permitting disputes, political opposition, and sustained community resistance. These developments follow reporting on heightened regulatory scrutiny in Virginia, where state officials revised proposed diesel generator permitting rules following months of public debate over emissions, noise, and community impacts.
Taken together, these Virginia projects, opposition data, developer surveys, and industry research suggest community acceptance is emerging alongside power availability, land, water, fiber connectivity, and utility infrastructure as a variable shaping where AI infrastructure gets built.
According to Matthew Shaw, the volunteer who compiles the opposition data, the methodology tracks only local Facebook groups dedicated to opposing data center development, excluding statewide and national organizations and petitions to reduce double counting. Shaw acknowledged that some individuals likely belong to multiple groups, meaning the data may both overcount and undercount the movement's size. "There is the issue of overlap," Shaw said. "We have no way of removing an individual who is part of multiple groups."
The fastest growth in opposition groups is occurring in Ohio, Pennsylvania, Michigan, Indiana, Georgia, and Texas—reflecting the geographic expansion of hyperscale development beyond traditional markets such as Northern Virginia. Shaw attributed the expansion of AI infrastructure and increasingly larger campuses as key drivers of organized opposition. While some overlap exists with the AI safety movement, most grassroots groups focus on physical effects of large-scale development rather than the technology itself. "The reaction is usually about property values, increasing electricity bills, emissions, increasing health issues like asthma, climate change, using water in drought conditions, coolant contaminating water and similar community concerns," Shaw said.
Until recently, developers evaluated prospective campuses primarily through technical and economic factors including power availability, land, fiber connectivity, tax incentives, and utility infrastructure. As AI campuses grow larger and impacts become more visible, community acceptance is emerging as another consideration influencing whether projects advance from announcement to construction.
Neil Osnato, founder of Persistence Analytics Group, said developers are increasingly evaluating community acceptance alongside traditional site-selection criteria such as power availability, land, fiber connectivity, utility access, and water resources. "Community acceptance has become part of the infrastructure stack," Osnato said. "For AI data centers, local opposition can now affect entitlement durability, permitting timelines, water approvals, backup-generation limits, utility planning assumptions, financing risk, and whether announced demand is actually executable at the proposed site."
Bloom Energy's survey reinforces this assessment. More than a quarter of respondents said community scrutiny has worsened over the past six months, making it one of the fastest-growing challenges facing new data center development after rising construction costs. Osnato noted that the industry's traditional sequence of securing land and power before addressing community concerns is becoming less reliable as AI campuses grow larger and more resource-intensive. Developers increasingly need to assess political and community support earlier in the site-selection process to reduce permitting and litigation risk. "Announced demand is not executable demand if the community, entitlement path, and local political environment cannot support the project," he said.