Local opposition to data centres has become a material issue in America's AI infrastructure buildout, but American Commerce Review's newest primary research suggests the underlying picture is not simply anti-development.
Among 6,415 respondents, 68.8% said they would support a new large AI data centre within 25 miles of where they live. Seventeen-point-six per cent said no and 13.6% were unsure.
Support comes with expectations about local value
Asked what would most increase support, local jobs ranked first at 24.6%. Lower electricity bills followed at 20.4%, operator-funded grid upgrades at 18.1%, renewable generation at 12.2% and local tax revenue at 10.3%.
Only 5.7% selected nothing, suggesting that for most respondents the local bargain around infrastructure, employment and household costs can influence acceptance.
Who pays for the grid is the clearest result
Forty-one-point-five per cent say data-centre operators should definitely pay higher electricity or grid charges when their projects require major new infrastructure, while another 34.9% answer probably. Combined, 76.4% support higher operator charges to some degree.
The survey does not determine the economically optimal tariff design. It does show that the cost allocation question is likely to remain central as utilities and regulators respond to unusually large new loads.
The result belongs inside the state-level AI race
ACR's state-level AI research tracks Texas, Virginia and other markets because data-centre economics depend on deliverable power, transmission, local regulation and the surrounding industrial ecosystem. Community acceptance adds another location variable.
A state can offer land and incentives while still facing resistance if grid upgrades, water use or electricity costs are perceived to fall disproportionately on existing customers.
Methodology
American Commerce Review surveyed 6,415 members of its audience in the October 2026 research wave. The survey is not weighted to represent the US population. Geographic subsets should be reported with their own sample sizes rather than treated as national polling.
| Measure | Share |
|---|---|
| Support large AI data centre within 25 miles | 68.8% |
| Oppose | 17.6% |
| Unsure | 13.6% |
| Operators definitely/probably should pay higher grid charges | 76.4% |
| Local jobs most likely to increase support | 24.6% |
| Lower electricity bills most likely to increase support | 20.4% |
| Operator-funded grid upgrades most likely to increase support | 18.1% |
Frequently asked questions
Is this nationally representative polling?
No. It is an American Commerce Review audience survey of 6,415 respondents. It should not be presented as the percentage of all Americans who hold these views.