Jessica Bichler is running for North Carolina state Senate this fall. The district where she's campaigning, in and around Wilmington, is deep purple, and she knew it was going to be a tough fight.
But even she was surprised when she appeared in an ad from her opponent, Republican incumbent state Sen. Michael Lee. The ad showed an AI-generated Bichler contorting her body into bizarre shapes, while making claims about her stances on policing and healthcare.
"I feel like I kind of hardened myself for some teasing and things like that, but the AI kind of took me aback," the real-life Bichler, who also owns a yoga studio, told NPR. "It was my face, it was my body moving around, but it was not anything I'd ever consented to or participated in creating."
The ad did include a disclaimer that it used AI, and in a statement, Lee told NPR that voters in his district "are smart enough to spot comedy and obvious parody when they see it."
"For years, leftist groups have distorted my image in ads. I didn't cry foul or complain; I worked harder to earn this community's trust," he added.
Deepfake candidates
Bichler is hardly the only candidate being rendered by artificial intelligence. New data shows AI political advertisements like these are popping up all over the country.
"It's getting way more common," said Michael Franz, co-director of the Wesleyan Media Project, which tracks political ad spending. Earlier this month, the group reported that so far this year, at least 164 political ads nationwide were produced or enhanced by AI. Some have been used in high-profile races, like Spencer Pratt's run for mayor of LA and the Senate race in Texas between James Talarico and Ken Paxton.
But many, like the ad against Bichler, are appearing in down-ballot or local elections. Franz said he's not surprised: "It's a way to enhance the content of your ads, make them look pretty sharp, and it doesn't cost you all that much."
AI-generated video was beginning to catch on during the last election cycle in 2024. At the time, many experts feared that it might lead to deepfake videos of candidates saying outrageous things. In response, 31 states have passed laws prohibiting deepfakes or requiring disclosures according to the National Conference of State Legislatures.
But Scott Babwah Brennen, director of the Center on Technology Policy at New York University, says the wave of fakes never arrived.
"The problem of deceptive deepfakes hasn't really materialized as a widescale problem," he said. "But something else is happening."
That "something" is more akin to satire: One ad shows Ohio Democratic U.S. Senate candidate Sherrod Brown overstaying his welcome at a birthday party. Another, just released by Texas Republican Senate candidate Paxton, shows his rival, Democrat Talarico, reading from a bible. Talarico responded by calling the AI ad deceptive.
Like the Bichler ad showing the yoga instructor in a physically impossible pose, these aren't meant to deceive. They're more like parodies aimed at making political points, or as Brennen puts it, "AI-generated slopaganda."
Just another tool
"The AI is simply a tool," said Todd Stiefel, a political independent who has funded primarily Democratic candidates in North Carolina over the years.
During the 2024 campaign, Stiefel created what is believed to be the first AI-generated advertisement in the state. It depicted then-Republican gubernatorial candidate Mark Robinson as an AI monstrosity with 14 fingers, while citing his alleged conspiratorial positions on a number of topics.
"We decided that AI could be a way that we could help people see these words that came out of his mouth, although technically they came out of his typing," Stiefel said. (Robinson subsequently lost the race, in part due to comments he'd reportedly made online).
Stiefel believes that AI can be seen as an extension of the use of art in politics, and others also see a connection. For hundreds of years, politicians have depicted opponents with exaggerated traits or shown them in ways that reflect their policies unfavorably, said Isabel Linzer, who tracks elections at the Center for Democracy and Technology, a think tank that follows AI's use in politics. In that sense, "AI is kind of facilitating the next version of the political cartoon."
But Linzer says these ads are also different — because AI technology can generate lifelike videos, and because it's so new, she thinks it could influence voters in really unpredictable ways.
"It's hard to make an assumption about how people who are scrolling online all over the country with all different kinds of backgrounds necessarily are perceiving the content that they're seeing," she said.
Did you consent?
Jessica Bichler said some of her supporters thought the yoga video generated by her opponent was real.
"People were actually questioning, 'Like, well how did he get into your studio space? How did you consent to be on camera for him?'" she told NPR in an interview. "I just had to explain 'Well, if you actually look at my legs and stuff, that's not actually humanly possible.'"
She sent a cease-and-desist letter to the Lee campaign about the ad, but it was over the way he represented her positions on issues, not his use of AI.
The Lee campaign responded by putting out a new ad, which toned down some of the claims and had a much larger AI disclaimer at the beginning. In his statement to NPR, Lee noted that he voluntarily put up the disclaimers on both versions of the ad because "AI imagery can be taken too far."
Bichler says she thinks AI's use in political ads is here to stay, but she doesn't plan on using it herself. She hopes to win instead, with what she calls "facts and reality."
WHQR's Ben Schachtman contributed to this story.
Contact Geoff Brumfiel on Signal at gbrumfiel.13
Copyright 2026 NPR
