Midterm Election Results Deal Major Blow to US AI Data Center Expansion as Voters and Lawmakers Push Back

Data center opposition had been building steadily across the United States, advancing one moratorium or blocked local project at a time. Then, virtually overnight, the midterm election fundamentally tilted a broad swath of the regulatory map against the aggressive artificial intelligence infrastructure buildout, catching investors and technology giants off guard.

Newly elected governors in Ohio, Georgia, Michigan, New Mexico, and Texas campaigned heavily on promises to stop—or at least temporarily pause—data center construction statewide. Meanwhile, incoming leaders in New York and Oregon pledged to prolong existing pauses already in place. Gubernatorial candidates calling for similar moratoriums on data centers also held leads in Alaska and Kansas, though those races remained uncalled as votes continued to be tallied.

All told, approximately 40% of pending data center capacity in the United States is planned for construction in states where newly elected governors have explicitly promised to halt or pause such projects, according to market intelligence data from Cleanview. This regulatory hurdle compounds local resistance, sitting on top of 46 local ballot measures that were either approved or well on track to pass, further stalling development according to Bloomberg Economics. At least 17 of those local measures represent outright bans on large-scale data facilities.

“It was obviously not a very good night if you’re a data center developer,” said Justin Teresi, a Bloomberg Intelligence policy analyst.

These political victories underscore a rapidly growing voter backlash against the massive industrial warehouses required to crunch complex queries for generative AI platforms like ChatGPT and Claude. The potential wave of statewide moratoriums and local restrictions directly threatens the development roadmaps of major tech giants who are frantically racing to secure additional computing capacity to meet soaring industry demands.

Financial markets reacted swiftly as investors braced for the fallout. Technology stocks plummeted following the election results in expectation of tighter regulatory scrutiny and operational hurdles. The S&P 500 Index dropped more than 4%, while the technology-centric Nasdaq 100 tumbled 6.4%, erasing a staggering $2.3 trillion in overall market value.

Since 2022, when OpenAI publicly released ChatGPT to the world, technology stocks have served as the primary engine behind an unprecedented bull run in US equities. Over the four-year period leading up to Wednesday’s market close, the S&P 500 had more than doubled in value, with the technology sector expanding to account for more than half of the benchmark index’s total value.

The election results add intense pressure to AI data center developers and their corporate customers, who were already navigating rising skepticism from investors and lenders. Financial institutions have grown increasingly wary of the heavy borrowing habits and the hundreds of billions of dollars being poured into the ongoing infrastructure boom. Local opposition efforts had already hampered at least 45 individual projects during the second quarter of 2026 alone, according to research group Data Center Watch.

Although data center oversight has increasingly evolved into a bipartisan issue, the election results mean that tough campaign rhetoric regarding AI infrastructure is highly likely to translate directly into state-level policy. Democrats have secured control of key branches of state government in several major data center hotbeds, including Virginia, Pennsylvania, and Arizona. According to a new report from BloombergNEF, these administrations are expected to pursue utility and permitting policies specifically designed to slow or halt construction.

Dan Diorio, executive vice president of state policy and government affairs for the Data Center Coalition, an industry alliance, warned that communities choosing to prohibit data centers risk watching valuable investments, high-paying jobs, and critical tax revenues migrate elsewhere.

“That’s going to go to other communities that have not shut themselves off,” Diorio said, “or to states that have not shut themselves off.”

Hyperscalers—the massive cloud-computing providers driving much of the expansion—are now expected to drastically overhaul and shift their development plans in response. Tuesday’s election results stand to shrink the $1.4 trillion that technology companies were collectively expected to spend on AI infrastructure in 2027 by roughly 5% to 7%, according to Mandeep Singh, a technology analyst at Bloomberg Intelligence. That anticipated reduction includes planned capital expenditures on advanced semiconductors, networking equipment, physical land acquisition, and other raw materials necessary for data center construction.

Those spending cuts and the anticipated slowdown in regulatory approvals could easily translate into as many as 10 gigawatts of planned data center capacity failing to come online next year as originally scheduled, Singh added.

“We’ve gone from the question of who is going to pass a moratorium to where can I even construct one of these things now?” Teresi noted.

Gordon Johnson of GLJ Research pointed out in a note to clients that Texas-based data center projects lacking pre-existing electricity allocations and environmental permits will likely need to look elsewhere. Such projects will “probably go to Louisiana, Oklahoma or nowhere,” he wrote.

From a purely political perspective, candidate promises to freeze the rapid AI buildout proved to be a popular strategy. Nearly three in four Americans now oppose the establishment of a data center in their local community, up significantly from about two-thirds in March, according to a September Quinnipiac poll. Furthermore, an October poll conducted by Ohio Northern University revealed that 57% of respondents in Ohio would vote “yes” on a total ban targeting large-scale data centers.

Nearly half of the nation’s 46 ballot measures aimed at restricting data center construction were concentrated in Ohio, where virtually all of them passed easily. Some of these measures targeted specific projects, such as EdgeConneX’s planned complex in Ashville, or imposed strict financial guarantees. Others established hard caps on the amount of power and water data facilities could consume, with 11 Ohio municipalities passing total or near-total bans.

Ohio has historically functioned as the “Wild West” of data center construction, according to David Niven, a political scientist at the University of Cincinnati. Bolstered by relatively inexpensive land and lucrative sales tax exemptions extending across decades, Ohio has amassed roughly 200 data centers, giving it the fifth-highest total in the country.

“The practical consequences of the backlash really only took off within the last few months,” Niven observed. “Data centers played a key part in the sort of general dissatisfaction that Ohio voters felt.”

While statewide executive actions carry significant weight, local ballot measures will exert a more direct and immediate effect on the industry because cities and townships maintain primary control over municipal zoning laws and property taxes, explained Joshua Basseches, an energy public policy expert at Case Western Reserve University in Cleveland.

Nevertheless, once Ohio Governor-elect Amy Acton assumes office in January, she will hold substantial executive authority. She can issue executive orders to halt state-level data center permitting or dial up regulatory scrutiny over their ongoing operations through targeted appointments to key state agencies, including the public utilities commission. Throughout her campaign, Acton frequently repeated the stump mantra: “Ohio is open for business, but we are not up for sale.”

Implementing more permanent structural changes, such as extending the temporary tax credit freeze initiated by her predecessor, will ultimately require direct legislative buy-in from lawmakers.

Even in states where outright moratoriums or comprehensive bans are not formally enacted, the operating landscape for hyperscalers looks increasingly hostile given the extensive authority state and local governments wield over land-use decisions.

“We expect to see a lot more sand in the gears of the AI build-out regardless—with more time required to secure permits and other approvals, as well as more concessions that data center developers will have to make,” said Libby Cantrill, head of public policy at PIMCO.

The political shift extends far beyond Ohio. Incoming governors in Iowa and Nevada have pledged to eliminate tax breaks tailored for artificial intelligence projects. Meanwhile, Byron Donalds, the governor-elect of Florida, has promised to pursue legislation requiring data centers to independently source their own electricity rather than drawing power directly from the public grid.

Even politicians who initially championed the rapid expansion of artificial intelligence shifted toward more critical stances during the final stretch of the recent election cycle. Republican Vivek Ramaswamy, who challenged Acton in Ohio, floated a policy proposal that would have required data centers to supply free electricity to nearby local residents.

“If one gubernatorial candidate went further, the other would follow,” said Morgan Scarboro, vice president at MultiState, a state and local government consulting firm. “This is an issue where it’s very difficult to just look at a quote and know which party is speaking.”

Sarah Adkisson, an attorney at business advisory firm EisnerAmper, suggested that local and state governments may increasingly pivot toward a “prove it first, then you get it” framework, withholding tax incentives and financial breaks until developers fully deliver on promised jobs, local capital investment, and municipal revenues.

Sensing the shifting tide, major technology companies had already begun formulating strategies to repair relationships within the communities where they hope to expand. OpenAI announced that it will actively avoid using nondisclosure agreements when negotiating future data center deals and will take affirmative steps to ensure local residents do not bear the financial burden of the electricity required to power those massive projects.