A.I. Is Coming for the 2026 Midterms

The rise of artificial intelligence is rapidly reshaping how we live and work — and it has the potential to transform our politics.

To learn more, I spoke with Kevin Roose, a technology columnist for The New York Times, a co-host of our tech podcast “Hard Fork” and, as you will see below, an expert in making complex issues accessible and even entertaining.

Kevin chatted with me about A.I.’s impact on American politics and where the fight goes next. I learned a lot from our conversation and hope you do too. As always, this interview has been edited and condensed.

Kevin, thanks for doing this. Increasingly, I’ve noticed politicians talking about the need to address the rise of A.I. — but often they’re speaking in broad strokes. Can you walk us through the most important political fault lines on this issue?

There are basically three camps.

The first camp — you could call them the A.I. accelerationists — wants to let A.I. companies build as fast as possible, and get the government out of the way. This is the position of the Trump administration, and an increasing number of A.I. companies are getting behind it. Most notably, the OpenAI president, Greg Brockman, and other industry leaders are backing a super PAC called Leading the Future, which says it has raised more than $100 million to spend on the midterms.

The second camp is more worried, and they have a laundry list of concerns about issues like selling A.I. chips to China and rules around deepfakes and A.I. companions for kids. There is also growing alarm over the capability of the systems themselves. Anthropic, one of the leading A.I. companies, recently put $20 million into a super PAC to counter the OpenAI effort.

Then there’s what you could call the populist anti-A.I. camp, which is worried about things like widespread job loss from automation, and is rebelling against the construction of data centers and other A.I. infrastructure. These people see A.I. as another example of “big tech” intruding on their lives, and they’re using a similar playbook to the ones that have been used to take on social media companies.

Do you see these fights playing out along ideological lines — left, moderate, right — or is there the potential for interesting new coalitions?

These camps don’t map neatly onto political parties yet. There’s a group on the right (including Steve Bannon) who are very worried about A.I. risks, conservative China hawks who disagree with the White House’s position on chip controls, and progressives like Bernie Sanders who want to block the construction of data centers. But I think these issues are somewhat “pre-polarized,” and I wouldn’t be surprised to see interesting new coalitions emerge.

Do you have a sense yet of what the biggest tensions around A.I. policy may be by this fall, as the midterm elections arrive? Or even in the next presidential election?

The big one looks like it’s going to be data centers, which are the most visible symbol of the A.I. boom, and a thing many communities are already upset about. But I’d also bet that job displacement will be a big issue in 2026 and 2028.

We aren’t seeing widespread unemployment from A.I. in the economic data yet, but as these tools get more capable of doing entry-level white-collar work, I wouldn’t be surprised to see a growing resistance to their use inside workplaces as a substitute for human labor.

You mentioned data centers. For those who haven’t been following closely, what are those, and why are they controversial?

A data center is, basically, a giant warehouse full of computers that power A.I. systems. The thing that makes the centers controversial is the scale. The biggest data centers are the size of small towns, and hundreds of billions of dollars are being spent to make even bigger ones.

They use a lot of energy, and some communities are seeing their electricity bills spike as a result of the increased demand on the grid. They’re also just a visible symbol of the A.I. boom, and a lot of people don’t want them in their backyards.

Is there any politician who stands out for discussing — or acting on — A.I. issues in an especially savvy way?

Not really! Most national politicians are still finding their feet on these topics. (That’s probably too generous.)

But I’m interested in what’s happening with Alex Bores, a Democratic state assemblyman from New York who’s running for Congress. He’s been a vocal supporter of A.I. regulation, and his race has become the first major target of the A.I. industry’s political machine. (Leading the Future is spending heavily on attack ads against him, and a super PAC affiliated with Anthropic is running ads for him.)

It’s the first real test case for how this stuff will play in a national race.

As you noted, leading figures in the A.I. industry could spend a lot in the midterm elections. What is it they want from the candidates they’re endorsing?

The accelerationists want America to “lead” on A.I., which usually means rolling back restrictions on A.I. developers and overhauling the permitting process to make it easier to build data centers. It could even mean supporting a plan known as “federal pre-emption,” which would mean imposing a multiyear moratorium on state A.I. regulations, so that, for example, New York and California couldn’t pass stronger regulations than the rest of the country.

The more worried crowd’s position boils down to, “We should have more control over this technology, and visibility into how it’s being built.”

What else is important for us to understand about how A.I. debates could shape our politics in coming years?

The rise of crypto’s influence in politics is instructive here. A few years ago, barely anyone in Washington cared about crypto, and the leading crypto companies could hardly get meetings with members of Congress. Then the money started flowing, and all of a sudden there’s a well-funded crypto lobbying apparatus in Washington, and not much opposition to it.

But I think A.I. is a bigger deal than crypto, and its effects are already more widespread. Americans are very nervous about the increasing role that A.I. is playing in their jobs, their kids’ schools, and the potential it has to transform their lives.

There’s going to be a lot of energy there, for whichever group figures out how to harness it.

In the spirit of A.I. discussion, I asked ChatGPT to name the single biggest political question about the future of A.I. It replied, “Who controls A.I. — and in whose interests does it operate?” Thoughts?

This is the big question about A.I. as it exists right now, but the people I talk to in San Francisco’s tech industry are more worried about where A.I. is going, and whether anyone can control A.I. systems at all. They see how quickly these systems are improving, and they’re starting to plan for what happens if and when they become smarter than humans, and capable of acting autonomously in the world.

If that happens, we’re going to have a whole new set of questions about A.I. politics, and it’s going to get much weirder than arguing about data centers.

The Supreme Court declares independence

President Trump has had an extraordinarily successful run before the Supreme Court. That came to a jolting halt with the tariffs ruling, my colleague Adam Liptak writes.

One Number

61 percent

That’s the share of Americans with unfavorable views of the Democratic Party, according to Gallup . Ruth Igielnik, The Times’s polling editor, explains.

This number is up from the 55 percent of Americans who viewed the party unfavorably in 2024.

Negative views of both parties have risen over the last few decades, but Democrats, in particular, have been viewed more negatively in recent years.

That is largely because rank-and-file Democrats are unhappy with their party. Just 70 percent of Democrats now say they view their party favorably, down from 85 percent in 2024, according to polling from the AP and NORC.

The Trump administration’s very manly week

Some of the Trump administration’s top officials want Americans to know one important thing: They are very manly men.

Defense Secretary Pete Hegseth bench-pressed more than 300 pounds. Health Secretary Robert F. Kennedy Jr. — shirtless and wearing jeans — worked out with Kid Rock.

My colleague Katie Rogers took stock of this hypermasculine social-media display and, she writes, “this administration’s hard-line approach to maleness.”

TAKE OUR QUIZ

This question comes from a recent article in The Times. Click an answer to see if you’re right. (The link will be free.)

Why did President Trump call himself a “schmuck” during a speech in Georgia?

Ruth Igielnik, Taylor Robinson and Ama Sarpomaa contributed reporting.

Katie Glueck is a Times national political reporter.

The post A.I. Is Coming for the 2026 Midterms appeared first on New York Times.