With midterm elections five months away and AI policy promising to be a political flashpoint, that map could come in handy for Democratic candidates—either as an asset, or as warning of their vulnerabilities. 

AI anxiety has escalated to a national conversation, but actual exposure to the technology remains relatively confined to specific cities and states. Workers whose roles are most exposed to AI are disproportionately clustered in Democratic-leaning jurisdictions, according to an analysis by the Brookings Institution published Wednesday. For a party still struggling to revitalize itself after widespread electoral losses in 2024, AI fears could emerge a make-or-break factor this November.

In 2024, 62 of the 100 most AI-exposed counties voted Democrat, according to Brookings, which defined counties as AI-exposed if larger shares of their workforces performed roles that could be handled by AI. Those places include traditional blue strongholds like Manhattan, the Bay Area, and Seattle’s King County, though several swing states—all of which were won by President Donald Trump in 2024—also reflect high degrees of AI exposure, including Arizona and Georgia.

The growing anxiety and public backlash over AI is likely to play a significant role during the midterms. Polling this year suggests a growing share of Americans are more likely to have negative feelings toward AI than positive ones, and a majority consider AI’s deployment to come with significant risks. A survey last month by researchers at the University of Pennsylvania found 65% of Americans say the government is doing “too little” to regulate AI.

“The ‘techlash’ against artificial intelligence is spreading, driven by workers’ fears of the technology’s potential to disrupt their jobs and upend their livelihoods,” the Brookings authors wrote. 

Some Democrats have seized on the issue and made it central to their campaign strategy, with some legislators calling for moratoriums on data center development and criticizing the Trump administration for its AI approach. The messaging might pay off, but given that the bulk of AI anxiety is concentrated in places that already voted blue two years ago, Democrats also stand to lose the most should AI backlash turn local.

A double-edged map

It isn’t necessarily a surprise the country’s AI-exposed geography currently leans Democrat. The Brookings authors called this a case of “occupational sorting.” Democrats tend to govern in urban areas where white-collar and office work represent bigger proportions of the local economy, performing the kinds of roles firms are experimenting with AI for. The Brookings report described Democrat-run, high-exposure counties as “hotbeds for some of the AI era’s most agitated voters.” 

This concentration of AI exposure could be a built-in advantage for Democrats, at least while Republicans control all three branches of government. Rising electricity prices, in part driven by data center energy needs, have already featured prominently in campaigns in which Democratic candidates have appealed to voters’ affordability concerns, and lawmakers including Sen. Bernie Sanders (I-Vt.) and state Sen. Kristen Gonzalez (D-N.Y.), have publicly backed plans to institute moratoriums on data center construction.

It’s what Rebecca Lissner, a former Biden administration advisor on national security, recently referred to as the “populist AI backlash.” In an article for Time in February, she wrote that the brewing AI discontent among many Americans could eventually unfurl into a “potent populist political force” for either party to take advantage of.

Beyond blue strongholds, Democrats could find success with their AI messaging too. Democratic candidates, such as Michigan’s Mallory McMorrow, have recently put out their own AI action plans covering online safety and tax policies to govern AI-driven wealth gains. Democrats also might hope their messaging can land with audiences across the political spectrum. The recent University of Pennsylvania survey found 53% of Republicans say the government should do more to regulate AI, as well as 73% of independents—roughly the same share as Democrat respondents. 

Data center construction has been primarily popping up in rural states, including swing states such as Arizona, Georgia, and Nevada. Local activists have also made Texas—where Democrats hope candidate James Talarico can flip a Senate seat blue in November—a hotbed for anti-data center protests. Talarico has consequently used his campaign trail to call for stricter regulations on data centers’ environmental impact.

Yet the same geography that makes AI exposure a potential Democratic asset could turn it into a liability. Despite the AI backlash and Trump’s own plummeting polling, Democrats remain uniquely unpopular too. A March NBC News poll found public opinion on AI had fallen off a cliff, though Americans were even less favorable of the Democratic Party. The same poll found while only 20% of Americans say Republicans would do a better job at governing AI, trust in Democrats was even lower, at 19%.

Democrats have been scrambling to nail down their message on AI for months, banking on riding affordability concerns and the AI backlash to a strong showing in November. But should their message not land, they risk not only a failure to get independents on their side: They might lose trust among the majority of AI-exposed Americans who already voted for them.