In the just-concluded 2026 U.S. midterm elections, the development of artificial intelligence and data centers encountered unprecedented public resistance. As voters overwhelmingly backed candidates and ballot measures that would restrict AI and data centers, the previously surging U.S. AI infrastructure buildout has run into a broad and major slowdown.
Local Referendums and Governor Races: Data Center Expansion Takes a Heavy Hit
According to statistics from polling firm Redbud Consulting, of the 44 proposals introduced in 10 states aimed at restricting data centers, as many as 38 have now been successfully passed, including dozens of moratoriums and bans. These policies are expected to have far-reaching effects on future projects planned for 2028.
In key governor races, several candidates who advocated strict limits on AI construction won decisive victories. For example, Democrat Gina Hinojosa won the Texas governor’s race. As the center of the U.S. AI boom, Texas is expected to account for more than 41% of new data center capacity in 2028. With her election, about 13 gigawatts of local data center capacity will face delayed review. Hinojosa plans to postpone related projects until new laws are passed to fully address the negative effects of data centers on water, electricity, air, and property rights.
Beyond Texas, Democrats also won Arizona, Georgia, and Ohio—three other major states that have been actively courting data center developers. Industry analysts predict that by 2028, nearly two-thirds of data center capacity may face delays of varying degrees, and half of U.S. data centers are now more vulnerable to strict scrutiny.
Tech Giants Downplay Results as Civil Rights Groups Point to Voter Backlash
In the face of the election results, the tech industry quickly offered interpretations and responses. Adam Kovacevich, a former Google executive and CEO of the tech industry trade group Chamber of Progress, publicly played down the results, arguing that this was not a collective rejection of technology or AI, but rather voters’ demand for economic change. He bluntly attributed the controversy to an “anti-Trump wave, not an anti-tech mandate.”
However, Alejandra Montoya-Boyer, vice president of the Leadership Conference on Civil and Human Rights’ Center for Civil Rights and Technology, took the opposite view. She noted that NBC News exit polls showed that as many as three-quarters of voters opposed data centers, and more than half of voters said AI and data centers were a key or important factor in their vote for Congress. This large-scale backlash stems not only from direct harm to local interests, such as rising electricity bills, but also from deep public distrust of Big Tech’s unchecked growth and regulatory failure.
Musk Bears the Brunt as Global Industry Landscape Quietly Shifts
In this election contest, billionaire Elon Musk became a major focus of attention. As a donor who poured large sums into Republican candidates and AI political action committees, Musk is currently facing litigation over pollution from gas turbines used at the Colossus data center in Memphis, Tennessee. Despite numerous political obstacles and lawsuits, Musk has not slowed down. According to SemiAnalysis, his SpaceX has even raised its 2027 data center capacity target from 10 gigawatts to about 13 gigawatts.
At the same time, the collective slowdown in U.S. domestic AI infrastructure construction has drawn widespread attention from international economists. Gary Ng, a senior economist at Natixis, pointed out that the slowdown in U.S. computing capacity construction and policy restrictions could keep U.S. AI token prices high, which objectively may provide more opportunities for overseas markets such as China to export cost-effective inference services. If opposition within the U.S. continues to grow, it is possible that some tech giants will be forced to move hyperscale data centers overseas, bringing new variables to the global AI industry and geopolitical competition landscape.
