Deepfakes hit East Valley campaigns
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Sarah Grillo/Axios. Stock: Getty Images
AI-generated images are becoming a flash point in East Valley elections, with Mesa and Scottsdale political figures accusing campaigns and outside groups of circulating fabricated visuals.
Why it matters: Arizona's new deepfake laws face an early test as campaigns use AI-generated videos and photos faster than voters, platforms and candidates can verify them.
Driving the news: Democratic state Rep. Lorena Austin, who's seeking reelection in Mesa's Legislative District 9, is considering legal action after two Turning Point Action activists posted images falsely depicting Austin campaigning for Republican gubernatorial nominee Andy Biggs.
- The images showed someone resembling Austin planting pro-Biggs signs in Mesa.
- Austin said they support Democratic Gov. Katie Hobbs, not Biggs.
- Mismatched tattoos, a watch that changed wrists and inconsistent language on the sign provided clues that the images were fake, according to KJZZ.
What they're saying: Mary Ann Mendoza, one of the Turning Point activists who posted the image on X, told The Arizona Republic it was a joke whose "humorous intent" should've been obvious.
- The now-deleted posts by Mendoza and Matt Whitmire had racked up roughly 25,000 views.
- Biggs spokesperson Drew Sexton told the Republic that the posts weren't connected to the campaign.
In Scottsdale, Council member Adam Kwasman accused council candidate Raoul Zubia's campaign Thursday of using a deepfake image of him in a campaign video, saying in a post on X: "This is simply shameful. You're lying to Scottsdale residents."
- Zubia's ad depicted an image of Kwasman for less than two seconds. It was edited out of the video less than 12 hours after it was launched.
Yes, but: Zubia spokesperson Matt Grodsky tells Axios that the campaign's video editors pulled the image from a May Scottsdale Progress article without realizing it was AI-generated.
- "Once it was determined to be an AI-generated image, it was removed within hours and replaced," Grodsky says.
State of play: Arizona law bars anyone from creating deepfakes within 90 days of an election.
- Anyone who creates and distributes a deceptive deepfake of a candidate is required to clearly label and disclose it as AI generated.
- The law requires proof that the creator intended to harm the candidate's reputation and mislead a reasonable person, and exempts satire and parody.
- A separate statute lets candidates and other Arizona residents ask a court to declare an image a digital impersonation.
The big picture: AI can degrade public trust even when people aren't fooled.
- "The main harm from the rise of AI-constructed media is not that people will be duped by something that is AI generated," Nate Persily, co-director of the Stanford Law AI Initiative, tells Axios. The bigger worry, he says, is that bad actors can dismiss something real as another deepfake.
- Even when voters recognize an ad as synthetic, the emotional impact can persist, UC Irvine epistemologist Cailin O'Connor says: "If you see some public figure saying something bad or doing something bad, you might still have negative emotions about them as a result."