David’s view
Republicans are very comfortable using AI for almost anything in politics, and they’ll say so. Democrats are not very comfortable, and they won’t talk about it.
We’re only starting to see how those different views might change campaigning. But Republicans are optimistic that their approach will win out.
“Professionally we have an advantage, because we’re willing to push the envelope on this stuff and take risks,” said GOP strategist Eric Wilson, an AI evangelist. “The professional class on the Republican side is very much saying: ‘Yeah, let’s use everything we have to our advantage. There’s no prize for running an artisanal campaign.’”
In Democratic shops, the conversation about whether to use AI tools internally is quiet and cautious. One example: I’m told the Democratic National Committee has approved the use of only Google’s Gemini LLM, and not its competitors, for official work like data analysis and coding. Other Democratic groups have used AI tools to build campaign sites more quickly, replacing some hours of busywork.
But there is Democratic dread about the consequences of widespread AI use, like employment loss for visual artists and camera crews, that the GOP doesn’t share. While data center construction and favorable tax subsidies for the AI industry are politically unpopular, only Democrats have a faction that wants to stop the bulldozers. Democrats have introduced the sole successful proposals to restrict AI in political messaging.
Rep. Julie Johnson, D-Texas, author of a bill that would let candidates sue for damages if their opponents ran “materially deceptive” AI ads, told me that such deception was already a problem in her primary.
“It depends on if it’s based in fact, or if it’s just made up bullsh*t,” she said, a few days before being forced into a runoff with former Rep. Colin Allred. “Colin has done some AI-generated ads with my face in them; I don’t like it, but I don’t have a huge problem with it. That is, in fact, me. What I don’t like is using AI to create an image of me with Trump and Kristi Noem, because that’s just never happened.”
That defense of observed reality doesn’t have much power on the right, which has fully embraced AI image generation. The tears of liberal-leaning artists who won’t get paid to make those images are a bonus.
And it’s not just the Trump operation diving in. The president’s amplification of AI images has often happened in a political vacuum — replies on X, re-Truths on the president’s social media account. By contrast, the National Republican Senatorial Committee has used AI in ways that deeply anger Democrats. Not that they can do anything about it.
This cycle’s aggressive NRSC AI use started in October, when Senate Democratic Leader Chuck Schumer told Punchbowl News that his party was politically winning the government shutdown and that “every day gets better for us.” But he didn’t say it on camera — a problem the NRSC solved with AI, creating a Schumer simulacrum that smirked out the comment.
The backlash was brief and minor. Since then, the NRSC has used AI video generation to put leading Democrats in embarrassing situations.
Texas’ James Talarico posted, but did not say, that “radicalized white men” are the biggest threat to America. The NRSC made him say so, in a video.
North Carolina’s Roy Cooper did not pose outside of a women’s bathroom with a purple-haired bearded man, but he did in an NRSC clip, which had some fun with the AI disclaimer: “AI Slop, obviously Roy Cooper has much tinier hands.”
After a while, Democrats stopped complaining about it.
“Embracing AI has allowed the NRSC to work harder and smarter while being better stewards of donor dollars,” said NRSC communications director Joanna Rodriguez. “Committees and campaigns still need smart, talented operatives, but AI maximizes that same operative’s abilities in a way that can save thousands of dollars and hours, if not days, of work.”
Republicans can say that because they’re confident that Democrats won’t mobilize and match them ad for ad. There is too much social risk on the left in becoming the kind of campaign that uses AI imagery; “slop” causes a backlash among progressives, but not among conservatives. Democrats also run the risk of hypocrisy. If a candidate said anything skeptical about AI’s threat to workers or the proliferation of data centers, there’s a genuine downside to using AI tools to clean up a photo.
Is there a corresponding benefit to boycotting it – to running what Wilson called an “artisanal” campaign? There is, in theory, a large bloc of voters who fear AI’s implications, bigger than the one that celebrates them. Democrats are aligning themselves with that mindset, handing Republicans the mantle of the AI-image party … and waiting for the backlash.
In this article:
Room for Disagreement
It’s no coincidence that California Gov. Gavin Newsom’s busy social media accounts make more use of AI imagery than any other Democratic campaign. He is parodying Trump, and a movement that’s been using AI to instantly heat up memes, for years.
But his team isn’t the only Democratic one that thinks the party needs to get over its allergy to AI.
“Democrats should be using AI in ad-making,” said Julian Mulvey, a Democratic ad maker at Fight Strategies, the firm that makes populist content for Zohran Mamdani and Graham Platner. “California has regulations so you cannot put words into opponents’ mouths, and there are quite rigorous disclaimer requirements. That’s good. [But] AI opens up too many visual storytelling opportunities to not embrace.”




