Political ads with made-up trans people are being used to bash Democrats
Elon Musk’s super PAC dropped dozens of ads this week in key Senate races (including in Ohio), and the majority of them focus on trans people. Others are resorting to AI.

This article originally appeared in The 19th and is reprinted here with permission.
By Brooke Migdon & Jasmine Mithani
America PAC, Elon Musk’s super PAC supporting conservative candidates and causes, is churning out dozens of digital ads vilifying transgender people and criticizing Democratic candidates over their past support for trans rights in the hopes of driving Republicans out to the polls in November.
In less than a week, the group, initially created to support President Donald Trump’s 2024 reelection campaign, has released at least 38 social media ads claiming Democrats in competitive midterm Senate races would “put our daughters in danger” and prioritize a “radical trans agenda” if elected. Most of the group’s ads focus exclusively on trans people, according to a review of America PAC’s ads in the Meta and Google libraries.
Two of the ads feature people who are implied to be trans women; it is unclear if they are real actors or AI-generated amalgamations. The ads were briefly removed for an unspecified “policy violation” after The 19th reached out to Google about whether they violated the company’s mandatory disclosure policy for AI-generated imagery in political ads. Two additional ads were also taken down. Google did not respond to another inquiry about which policy these ads violated. As of publication, the ads had been reinstated.
Musk’s super PAC is far from the only Republican-aligned organization to target the Democratic Party’s once-broad support for trans people ahead of the midterms, but it is one of the most powerful and well-funded. Musk, the world’s richest person, plans to spend at least $100 million to help elect Republicans this fall, the New York Times reported in July.
The ads, which portray trans women as drag queens rather than as they are in day-to-day life, follow a recent pattern of AI-generated attack ads targeting Democratic candidates over their support for trans rights. It’s unclear whether these specific ads are AI-generated, but they fall in the uncanny valley of authenticity. At minimum, they present harmful stereotypes of trans women — that they are physically imposing and always wear overdone makeup — to an audience who may not have ever met a trans person.
In an ad against Democrat Sherrod Brown in Ohio, an image of a crossing guard with a blowout and dark eyeshadow, implied to be a trans woman, looms in the background as a voiceover declares that Brown “gave the green light to the radical trans agenda” while in office. Brown was ousted from the Senate in 2024, losing reelection to Republican Bernie Moreno, and is attempting a comeback against Sen. Jon Husted.
“Tired of grown men dunking on female athletes?” a narrator, referring to trans women, asks in another ad targeting James Talarico, the Democratic Texas Senate nominee. A person who is implied to be a trans woman, wearing a sequined, rainbow-colored jersey with a face of heavy makeup and lacquered hair, slam-dunks a basketball in front of a group of young girls.
“They are trying to create a caricature of trans people that, to your average person, appears outlandish,” said Kelly Dittmar, a professor of political science at Rutgers University. “AI allows for literally creating the image of a person who doesn’t exist, and it allows them to do so in an easier way.”
In this new age of AI, it’s increasingly difficult to tell what is real and what isn’t — but experts say it doesn’t necessarily matter to the voters who see this kind of media. (America PAC did not respond to questions about whether the recent anti-trans ads were made using AI, and The 19th was unable to confirm whether the ads were synthetic.)
AI-generated content has been especially prevalent in Texas’s Senate race, where deepfakes of Talarico focus largely on his past comments about gender and trans youth. Talarico’s opponents have used AI to put him in a dress while he sings about gender-affirming care; dress him as a surgeon threatening a little girl on an operating table while a voiceover talks about “sex-change surgeries”; and, this week, show Talarico, a religious scholar, defacing a Bible.
That most recent ad comes directly from the campaign of Ken Paxton, Talarico’s Republican challenger and current Texas attorney general. Paxton has previously declared that Talarico is too “low-T for Texas,” implying a testosterone deficiency.
Talarico walked back some of his previous statements on trans people in a Fox News appearance on Wednesday: He said that there are two genders instead of six, as he said in 2021, and that he will “oppose trans athletes playing in girls’ sports if it risks safety or fairness.”
The Democratic nominee on Wednesday also shared his thoughts on Paxton’s latest AI ad, condemning it and accusing his opponent of being bought by the data center industry. “They are using AI in order to mislead the people of this state, and that is completely consistent with a liar like Ken Paxton,” he said.
Talarico told CBS News that he would not use AI in political advertising: “Every ad we put out is made by human beings, and we’re very proud of that fact.”
Experts say images, even fake ones, can evoke a strong emotional reaction.
Anti-trans campaigns, particularly related to sports, are often theoretical. Lawmakers sponsoring statewide bans on trans student-athletes, for instance, often struggle to cite local examples of trans girls playing on girls’ sports teams.
“The over-the-top nature of this stuff, at some point, is going to wear out its welcome,” said Lanae Erickson, senior vice president at Third Way, a left-of-center public policy think tank in Washington, D.C.
“When you talk about transgender women in sports, people already have a vision in their head of what that looks like, but they also know the difference between a grown man and a transgender girl,” Erickson said, referencing the America PAC ads. “The more you’re catastrophizing it and showing pictures of grown men, it’s just taking it so far that people may just turn off. There’s a point at which you lose people.”
AI makes it easier to convince audiences that trans people look a certain way, are trying to do a certain thing or push a particular agenda, said Blair Attard-Frost, a professor of political science at the University of Alberta who studies AI policies.
“All of the messaging and narratives they’ve created, all the trans panic they’ve created over the years, they now have this tool available to them that can create imagery that affirms that reality,” she said.
Laws on AI use in campaigns lag behind technological advancements. No unifying federal guidelines exist; at least 30 states have enacted some sort of law governing the use of AI in political ads, according to the consumer advocacy nonprofit Public Citizen, but they vary widely.
In Texas, deepfakes of candidates are banned only 30 days before an election, and they must have been published with “intent to injure a candidate or influence the result of an election.” Violating the law, passed in 2019, could result in a misdemeanor. But the Houston Chronicle points out that this only applies to state races, not federal ones, and that it hasn’t seen much, if any, use.
Ohio, where Brown is running, has yet to pass any legislation on the topic, but Public Citizen lists five attempts since 2024. The most recent bill was introduced on September 2 by Republican James M. Hoops; previous attempts have come from both parties.
This Trump administration is the first presidency to embrace AI-generated content in its political messaging, which set the stage for how the technology can be used — or misused — in elections, Attard-Frost said. She thinks the cozy relationship between Big Tech companies and the White House has facilitated that, in some ways.
AI-generated political content isn’t just pushed by those in office, and Attard-Frost thinks the broader normalization of “slop” will impact voters between election cycles. She’s concerned that AI-generated content from sources like PACs or elected officials will end up being more impactful because it reinforces what people already see online. “It kind of lends authority to this fake narrative that they’ve just seen playing out on social media,” she said.
And the impact of ads like the ones put out by America PAC, regardless of provenance? “The only way I see it going is more anti-trans sentiment,” she said. 🔥
Know an LGBTQ+ Ohio story we should cover? TELL US!
Submit a story!
