Curtis Hier uses “Planet Hank” as a verb. 

He used the controversial Brattleboro content creator’s moniker to describe the inspiration behind Flip Vermont Red, the political action committee dedicated to electing more Republicans to the State House that he founded last month.

“Somebody just sort of said to me … ‘Oh, I wish we could “Planet Hank” Dave Wolk,’” Hier said in an interview, referring to the independent State Senate candidate. “I started thinking I could ‘Planet Hank’ Dave Wolk.”

AI-generated political content, sometimes referred to as “slopaganda,” has gained traction on social media following a successful challenge of a state law regulating synthetic material during election season. Containing renderings of candidates and legislators, some experts and politicians are concerned that, despite the ridiculous and satirical nature of the content, slopaganda is changing the political atmosphere.

University of Vermont philosophy professor Randall Harp said he is concerned about the volume of AI-generated political content online. It can harm social and information systems, he explained. In an interview, he compared AI content to fashion magazines. 

“If we get flooded with manipulated images in certain spaces, like fashion magazines, … even though we all know or understand that these are manipulated images, it still has a cumulative effect on the way that we think about ourselves,” Harp said.

In the same way, he said, sometimes the human brain can’t process content as satire when it “looks like reality.”

Most political deepfakes — generally AI-generated images or videos that misrepresent a real person — don’t intend to mislead, but distinguishing satire from deception has proved legally intractable, researchers at Purdue University found.

Thirty-one states have enacted laws regulating the use of AI in political media. When Montana enacted its law requiring disclaimers on election deepfakes, Federal Judge Susan Watters struck it down, arguing that satire exemptions force government officials to make subjective judgments about protected speech. Vermont’s Act 75 contains the same exemption.

Flip Vermont Red’s first Facebook post was not AI-generated. It contained a real image of Sen. Ruth Hardy, D-Addison, overlaid with text criticizing her role in overriding Gov. Phil Scott’s 2024 veto on controversial land use bill Act 181.  

The PAC’s second Facebook video was AI-generated. It depicts a realistic rendering of Hardy nailing a “REGULATED BY THE STATE OF VERMONT” sign into a farm property. A farmer removes the sign from the dirt. Hardy then installs more signs, and hands the farmer a property tax bill. The farmer then packs boxes, farm animals and equipment into a moving truck. “AI made me do it,” small, white text in the video’s top left corner reads.

Hardy has been named or depicted in at least 11 posts by Flip Vermont Red.

Hank Poitras, known online as Planet Hank, has posted AI-generated political videos as recently as Sept. 22. 

A June 11 video Poitras posted depicted then-Republican U.S. House candidate Mark Coester criticizing U.S. Rep. Becca Balint, D-Vt., before using an excavator to lift her into a trash bin. 

Attorney General Charity Clark investigated that video for violation of Act 75, which requires that “realistic” AI-generated content contain a specific disclaimer. Poitras’ video did not contain such a disclaimer, but a federal judge ruled that the video fell into the law’s satire exemption.

Political satire and parody have deep roots in American discourse. First Amendment scholars note that exaggeration, irony, and invented scenarios are traditional tools of political critique. Hany Farid at UC Berkeley said deepfakes are becoming “more sophisticated and compelling,” but the research offers no clear path to regulation that survives constitutional scrutiny.

When asked for comment on the videos he had posted, Poitras wrote, “Planet Hank believes in and fully supports the First Amendment.”

Flip Vermont Red doesn’t include disclaimers in its posts, either. Hier said he is not concerned whether the PAC’s posts violate Act 75. 

“Our synthetic media are neither deceptive nor fraudulent,” Hier wrote in an email statement. “We use cartoons or outlandish imagery.”

Hier is a college application coach and a former high school teacher, according to his LinkedIn profile. He is also an administrator on the Facebook page Fair Haven, VT GOP, where he reshares Flip Vermont Red posts, uploads news articles and advertises events with the Rutland County Republican Office. 

He added in an interview that he aims for a 100% AI rating if his content were pasted into an AI-checker.

Hier said that “tossing Becca Balint into a dumpster is not something we would do,” because Flip Vermont Red’s message “is not ridiculous.” 

“There’s a point to it,” he said. “It’s not just, you know, take out the trash. It’s somebody is doing something wrong or that we believe is wrong, and we want to tell people about it.”

Hardy said that, while she has avoided most of Flip Vermont Red’s posts, she doesn’t believe they will impact her reelection campaign. 

“The people that I’ve heard from are upset by the videos,” she said. “They don’t feel like they’re appropriate. They don’t feel like they’re accurate. They don’t reflect me as a human, as a senator, as a policymaker, as their representative in the legislature.”

Wolk also wrote in an email that he believes in voters’ judgments when viewing this kind of content.

“They know the difference between my track record of public service in Vermont for over 50 years and an AI-generated portrayal from an anonymous account,” he wrote. “Too often the loudest voices on the left and the right consume all of the attention, while the important work facing Vermonters gets ignored.”

Hier provided VTDigger with an early cut of an AI-generated video which depicts a realistic Wolk in an office at Castleton State College. A cartoon Wolk then walks around the campus as buildings pop up behind him. He then sings a song, seemingly mocking Wolk’s spending during his 16-year tenure as president. Hier said Flip Vermont Red plans to use it in a Facebook ad campaign paid for by donors to the PAC. 

The video of Hardy will also be boosted as part of the advertisement campaign, he said. 

Hier uses a $9.99 monthly Adobe Firefly subscription to produce AI images. He said he outsources production of the videos that Flip Vermont Red posts, declining to comment on whom he outsources from. 

The lead sponsor of Act 75, Hardy believes it’s “no accident” that she is being targeted by AI-generated political content. She, now and in writing the bill, said she wants to protect the First Amendment and protect voters’ access to “accurate information.”

Hardy declined to comment on whether a court could find Flip Vermont Red in violation of Act 75, but said that the bipartisan-supported bill may need to be edited in the next session. 

“I think it sort of speaks to this moment that we’re in,” Hardy said. “The technology and the use of that technology is spreading, advancing and spreading so quickly that legislators can’t keep up with it.”