A new political ad put out by Attorney General Ken Paxton’s campaign claims his Democratic opponent in the U.S. Senate race, James Talarico, is weaponizing Christianity to sell a radical agenda.
The ad shows AI-generated images of someone who looks like Talarico sitting at a desk, writing and reading out his thoughts about gender identity and feminism in the Bible.
Rice University politics professor Mark Jones said the use of artificial intelligence in political advertising has grown exponentially over the last few election cycles.
“It’s a very effective way to do things that the campaigns used to do with voiceovers, reading what the candidate had said or written previously. Now instead of having a voiceover do it, you can have the candidate themselves say that via AI,” Jones said. “Of course, that wasn’t [actually Talarico], but it can be confusing for voters.”
Paxton is not the only one to use AI this way during this campaign season. Republican Lt. Gov. Dan Patrick ran an AI-generated ad attacking his Democratic opponent, Vikki Goodwin, alongside Talarico and Democratic gubernatorial candidate Gina Hinojosa.
“It was clearly AI, but it was done in a way that definitely had them saying things that they had not actually said, unlike the case of the Paxton ad where he’s quoting Talarico verbatim,” Jones said.
Texas was one of the first states to ban AI advertisements in campaigns back in 2019, Jones said, but only within 30 days of an election.
“There also are federal requirements that you disclose who’s doing the ad and in terms of funding. But in Texas, there is no restriction on these ads up until 30 days, prior to Election Day. So anybody can use them,” Jones said. “And even after the 30-day window has started, it’s not entirely clear if they apply to federal elections or not.”
And while both the Paxton and Patrick ads included disclosures saying they included AI-generated images, Jones said not every viewer will notice that.
“Not everyone’s going to be paying exact attention. And I think we’re going to continue to see these ads more and more, because especially if you look at the Talarico ad, it was a very effective way to get across the message,” Jones said. “But it can start to veer into more dangerous territory if we get into situations where the ads appear to be more realistic. Up until now, there’s so clearly AI inventions that it’s easy for someone to spot, but that doesn’t necessarily mean that’s always going to be the case.”
