AI Is Suddenly a Midterm Issue — and Voters Trust Neither Party With It

Artificial intelligence has completed a journey that took the internet decades: from techno-curiosity to midterm election battleground. On Tuesday morning, the New York Times published the latest Times/Siena national poll under a blunt headline — “How Popular Is A.I. With Voters? Not Very, a New Poll Finds” — and the numbers behind it sketch a political landscape where the most important technology of the era is also one of the least trusted, with no party having figured out how to own it.

The poll, conducted among registered voters with a margin of sampling error of plus or minus 2.9 percentage points, arrives at a moment when anxiety about AI has burst fully into the open. In the span of a few days, Anthropic CEO Dario Amodei published a 3,800-word essay calling for a global slowdown of frontier AI development, President Trump dismissed warnings from top technologists as a “HOAX,” Senator Bernie Sanders and Steve Bannon shared a stage at a pro-human rally in Washington, and Congress fought openly over guardrail legislation. The Times/Siena survey is the first major national measure of how all of that noise is actually landing with voters. The answer: badly, for everyone.

The One Issue Where Republicans Lead

The survey’s most striking finding is a negative image of the Republican Party’s broader predicament. Asked which party they trust more to handle artificial intelligence, voters gave Republicans a 42 percent to 40 percent edge — a statistical tie inside the margin of error, but notable for one reason: it is the only issue in the entire poll where the GOP holds any advantage.

Democrats lead on everything else, including territory Republicans held for years. On the economy, voters prefer Democrats by 7 points; on the cost of living, by 8; on immigration, by 7; on the deficit, by 10. Against that backdrop, the two-point AI edge is less a Republican strength than an artifact of Democratic incoherence — the polling equivalent of winning by default.

The demographic splits underneath are dramatic. Republicans lead on AI trust by 15 percentage points among voters aged 45 to 64, but trail Democrats by a staggering 32 points among the youngest cohort, ages 18 to 29. The gender gap runs the other direction: the GOP does better with men (plus 14) than with women (minus 6). AI trust, in other words, tracks the same fault lines of age and gender that now structure American politics almost everywhere else.

Neither Party Has a Position

Why can’t either party break open an issue where the public is this uneasy? Because, as the Times reporting makes clear, neither side actually has a stable position to sell.

President Trump has leaned into what pollsters describe as an “all-gas, no-brakes” approach — championing AI acceleration, staging calls with Nvidia CEO Jensen Huang, and raging online this week against the “doomers” urging caution. Some Republicans in Congress are more nuanced, pushing safety legislation even as the White House attacks the very premise of AI alarm. The result is a party simultaneously running as the party of AI optimism and the party whose own legislators are holding hearings about extinction risk.

Democrats face the mirror-image problem. The party is “all over the map” — talking up the urgency of AI risk without a unified vision of what to do about it, while trying not to antagonize a tech donor base and a labor coalition that fears job losses for entirely different reasons. When both offerings are this fuzzy, voters default to partisan identity, and the result is a 42-40 coin flip.

Voters Agree on One Thing: Not in My Backyard

If partisans can’t agree on who should steer AI, they converge sharply on the physical infrastructure behind it. The poll found voters united in opposition to the data centers sprouting across the country — the power-hungry facilities that turn abstract AI anxiety into concrete grievances about electricity bills, water use, and rural land. That bipartisan nimbyism has already produced strange coalitions in state legislatures, and it may prove more electorally potent than any argument about model capability.

This is not a new anxiety, and the trend line is consistent. An NBC News poll in March found a majority of voters saying the risks of AI outweigh the benefits, with voters trusting neither party on the technology. A Change Research survey in April found most Americans prefer stricter AI safety rules even if that means falling behind China. A Fox News poll in May found nearly 8 in 10 voters consider AI regulation urgent. The Times/Siena result confirms the pattern at the highest-profile polling operation in the country, seven weeks before Election Day.

What It Means for November 3

For candidates in competitive districts, the poll carries an uncomfortable message: AI is now salient enough to move voters, but the safest play may be attacking the industry’s footprint — data centers, energy costs, job displacement — rather than debating capabilities and safety thresholds that most voters cannot observe.

For the AI industry, the signal is worse. The companies racing to build frontier systems enjoy support from the White House, several governors, and a bipartisan chunk of Congress, yet the voting public is skeptical, frightened, and increasingly organized against the physical buildout. A political system that spends the next two years fighting over AI will not automatically fight for it.

And for the pollsters themselves, there is a historical echo worth noting: the internet went from novelty to political weapon in roughly a decade, and the first wave of regulation arrived only after the damage was done. AI has compressed that timeline into a single election cycle. On November 3, 2026, American voters will pass their first collective judgment on the machine age — and if this poll is right, they will do it holding their noses at everyone who claims to be in charge of it.