Widespread voter opposition to artificial intelligence (AI) data centers emerged as a pivotal force in the midterm elections, fueling Democratic gains and raising the prospect of stricter state and federal regulations on the technology sector, according to recent reporting and election analyses.

Across 10 states, voters approved restrictions or rejected data center proposals in 40 of 41 local ballot initiatives, according to tracking by Redbud Consulting.

The measures passed by an average margin of 49 percentage points, cutting across traditional partisan lines in communities ranging from conservative Edgewater, Fla., to progressive Alhambra, Calif. Only one restrictive measure failed, in Harwood, N.D., where local leaders successfully argued that an ongoing project was vital for municipal revenue.

The backlash against the massive energy- and water-intensive facilities also reverberated in high-profile statewide races. In Texas, where Bernstein estimates data center construction accounts for 28% of the state’s gross domestic product, Democratic Governor-elect Gina Hinojosa unseated long-standing Republican control after campaigning on a moratorium on new data centers until lawmakers pass consumer, union labor, and transparency protections.

Similarly, in Ohio, Democratic Governor-elect Amy Acton pledged moratoriums on projects failing to meet union hiring and industrial zoning standards, while Democratic Senator-elect Sherrod Brown targeted incumbent Jon Husted over his past support for the industry.

According to NBC exit poll data, 56% of voters — and 68% of Black voters — said data centers were a major or crucial factor in their congressional vote.

While Democrats accounted for roughly two-thirds of those voters, opposition spanned ideological divides, uniting environmentalists, farmers concerned about land use, and residents facing higher utility bills.

Ahead of the election, President Donald Trump pushed back against critics, arguing that opposing data centers would leave communities “backwards and poor.” Yet bipartisan concern in Congress had already prompted the House to pass the Ratepayer Protection Act by a 417-3 vote, though the bill stalled in the Senate over debates regarding voluntary versus mandatory utility rules.

News of the election results and the prospect of a Democrat-led Congress pursuing tighter oversight triggered an immediate sell-off in AI infrastructure stocks. However, Wall Street analysts and industry participants suggest the political shift will prompt a regulatory reset rather than halt the AI building boom.

Investment bank Bernstein projected that even in an extreme scenario, only 9% of planned data centers would fail to break ground, with a realistic cancellation rate of 2% to 4% nationwide.

Susquehanna Financial Group similarly forecasts that more AI-training data centers will be built in 2027 and 2028 than this year. Major tech companies, including Amazon.com Inc., Microsoft Corp., Alphabet Inc.’s Google, and Meta Platforms Inc., which are on track to spend over $720 billion on capital expenses this year, have deep financial reserves to absorb higher power costs, reduced tax breaks, and stricter regulations.

Microsoft, for instance, has already begun drafting model state legislation addressing tax revenue and water limits after internal polling showed 70% of residents would accept data centers with proper safeguards.

Smaller, heavily leveraged builders known as “neoclouds” face greater vulnerability to rising costs and delays, which former Microsoft executive Brian Janous said will “separate the wheat from the chaff.”

Meanwhile, political observers note that while exit polls do not fully capture the breadth of national resistance against more than 600 proposed facilities, the narrative of voter pushback is already reshaping economic expectations and setting the stage for data center regulation to become a central issue heading into the 2028 presidential primaries.