Democrats don’t yet have a consensus on AI. But blue states are the starting point.
By any reasonable measure, the political discussion around AI has fundamentally changed in the past few months — perhaps weeks. Polls show data centers’ popularity to be approaching the flu, and AI CEOs are even more unpopular than other corporate leaders. Safety concerns are jolting even once-skeptical observers. Candidates from both parties are rushing to recalibrate their messaging. Anecdotally, I haven’t heard this much proactive discussion of a policy issue among my friends and acquaintances in years. That President Trump remains unswervingly, almost comically, supportive of the industry would seem to present Democrats with a ripe political opportunity.
So far, that’s not the case. A New York Times/Siena poll last week, widely regarded as one of the best polls in the game right now, laid bare the party’s problem: The one policy issue on which Republicans held an advantage was AI.
Last week’s NYT/Siena poll.
Granted, the advantage was small, and the finding was an outlier compared to most surveys. But even more significant was the high degree of uncertainty in the electorate — more than any other issue, Americans did not know which party they trusted more to handle artificial intelligence. That, more than anything else, is the main finding of recent polls on the issue: Over the past month, multiple major surveys have found voters deeply uncertain as to which party they trust more on AI.
All four polls were released this month.
It’s in that void that some on the left are now making a new push for Democrats to claim the mantle of AI regulation — with some pointing to governors’ existing action as a model.
Skeptics’ growing momentum
Rishi Bharwani is the U.S. director of Reset Tech, a group advocating tech regulation. On Wednesday, the organization, in collaboration with Groundwork Collective, published a report outlining what it sees as the mounting economic threat from AI. The report argues the technology is already wreaking havoc, citing an uptick in online scams, the emergence of micro-targeted pricing, and rising electricity prices.
The group is focused on threats to consumers, rather than the generalized concerns about AI safety that have dominated the news in recent weeks. But in some ways, it’s responding to the same dynamics as so-called “AI doomers”: the technology has advanced so rapidly that the existing set of rules is inadequate for the new landscape. The lack of historical precedent is front of mind for the group. “The conditions are so ripe for exploitation,” Bharwani told me. “I really struggle to see an analog for something like this.”
Here it’s important to distinguish between Americans who follow the news closely and have likely seen evidence of Democrats’ sharp turn against the AI industry in recent months, and less-informed voters, with whom the contrast has yet to break through. It might not be long before it does: Across the country, the past few weeks have seen a flood of action around AI from state leaders, including JB Pritzker, Abigail Spanberger, and Gavin Newsom. With congressional Democrats handcuffed in the minority, Democratic governors’ recent actions could be the beginning of a long-term change in the party’s brand.
“This is a highly salient jump ball issue,” Bharwani said. “But the biggest thing that jumps out to me in the recent weeks with these governors taking meaningful policy steps forward is that the parties are in different places. Democrats have taken significant steps toward claiming that jump ball and Republicans have taken a few steps back — as long as Donald Trump and David Sacks continue to have a hands-off regulatory approach and a ‘let it rip’ approach to the deployment of AI and construction of data centers.”
The other side
Not everyone on the left is gung ho about becoming the anti-AI party. Leading the Future PAC, a big-spending super PAC working on behalf of AI companies’ interests, counts dozens of House Democrats as allies. Even Rep. Alexandria Ocasio-Cortez, who one might expect to be at the vanguard of anti-AI sentiment, has taken a conspicuously different tone on the technology than her political mentor, Sen. Bernie Sanders. Over the weekend, Ocasio-Cortez sent a letter to her email list that cast the leading AI companies less as existential threats and more as a new generation of plutocrats.
Koustubh Bagchi, vice president of U.S. Policy at the Chamber of Progress, a center-left group that takes a less restrictionist view of AI, is one voice urging Democrats to take a more measured tone. He singled out New York Gov. Kathy Hochul’s first-in-the-country data center moratorium. “That was sort of an outlier,” Bagchi told me. “That's not an approach we think is responsive to what voters want. If a community wants a data center, they should be able to work out a deal.”
Bringing up certain governors also exposes the gap. While Bagchi praised Colorado Gov. Jared Polis, perhaps the most pro-tech Democrat in the country, for taking a “more thoughtful” approach, Bharwani criticized him for being “clearly out of step” with other Democrats.
Interestingly, though, the actions of Democratic governors writ large so far unite both camps. New Jersey Gov. Mikie Sherrill was mentioned by both as a positive example for other Democrats, and recent actions taken by potential 2028ers were complimented. “You had New York and California set the standard for what frontier transparency legislation should look like,” Bagchi said. “Illinois added the compliance audit requirement, and then Massachusetts now is looking to build on that.”
At the same time, there’s reason to be skeptical that unity can last. One of the reasons state Democrats’ actions have not yet caused a real division is fairly simple: there’s only so much that can be done at the state level. As much as governors have an incentive to present themselves as taking aggressive action — a theme that came up in the above conversations and in discussions with a tech analyst who’s been in contact with advisers to potential 2028 candidates — they have not come close to achieving structural change to the industry’s trajectory.
“The real answer is where [Democrats] land the next time we’re in power,” that analyst told me. “And I’m not sure anyone knows yet.” For now, if the actions of the only Democrats in the country with any real power are indicative, the party's direction is clear, even if how far it will go isn't.
“It’s absolutely imperative that Democrats claim the upper hand on this issue,” Bharwani told me. “This is going to be a tier one issue in the presidential campaign.”
Worthy reads
Chris Lehmann of The Nation argues for Democrats to take advantage of Trump’s staunch AI support.
Daniel Donner of The Downballot finds one indication that 2026 could indeed be a blue tsunami: primary results.
Politico’s Jonathan Martin lays down a marker for the midterms: We’ll know a lot about how they’re going to go in about two weeks’ time.
Over at RealClearPolitics, Jonathan Draeger lays out what appear to be the best-case scenarios for both parties in November.
In the New York Times, Sabrina Tavernise casts 2026 as “the distrust election,” warning some voters now associate Trump with “the establishment he once railed against.”
And Adam Jentleson, the president of Searchlight Institute, presents his case for how to build long-term Democratic majorities.
Got feedback on today’s newsletter? Email me [email protected]