From a Kentucky congressional race to the Los Angeles mayor’s contest, deepfake political ads are turning up across the country. George Gerbo examines the science behind why they work, who’s most susceptible, and the bipartisan push — so far stalled — to rein them in.
This is the AI smear machine.
You may have seen political attack ads before, but probably not one like this:
Thomas Massie, caught in a throuple in Washington.
That’s from an attack ad targeting Kentucky Republican Congressman Thomas Massey. But the images in the ad aren’t real. They were made with artificial intelligence, showing Massey holding hands with Democratic representatives Alexandria Ocasio-Cortez and Ilhan Omar, and checking into a hotel.
This is worse than adultery. It’s a complete and total betrayal of President Trump and Kentucky conservatives.
Massey says the ad could mislead older voters, warning, “It reeks of desperation, but they’re hoping the older generation won’t realize it’s an AI-generated lie.”
Researchers who study aging and deception say he’s on to something. In this video, we’ll take a look at why deepfake political ads are fooling older Americans and what it means for the future of political campaigning.
Many of us probably have a parent or someone in our lives who we’ve needed to steer away from a suspicious email or a tempting link. The practice of phishing — a social engineering attack where scammers impersonate trusted entities via an email, text, or call to steal sensitive data — has been around since the advent of the internet.
Now, in an era where everything is politicized, new types of scammers, ones that are looking to change your thoughts, not your pocketbook, are using AI-created media to alter campaigns in real time.
National political reporter Seth McLaughlin talked with researchers at the University of Florida who say older adults face a unique set of challenges when confronted with AI-generated photos, video, and audio that look like the real thing.
Researchers have said part of the reason that works is that older voters in general are somewhat more trusting of what they read and see online. We’re starting to see these AI-generated ads proliferating across the country. - Seth McLaughlin
The Massey ad, funded by the MAGA Kentucky Political Action Committee, is a prime example. A small disclaimer notes the spot was created with AI, but the visuals do the work. Those visuals are typically called deepfakes, defined by the Government Accountability Office as hyper-realistic digital forgeries to depict real people doing or saying things they never did.
If that definition sounds overly wordy, you’re not alone. A 2023 Pew Research survey found nearly 6 in 10 Americans aren’t sure what a deepfake even is.
The ad, in part, did its job. After the most expensive race in terms of advertising in House history, Massey lost his re-election bid to challenger Ed Gallrein by 10 percentage points.
His opponent, Ed Gallrein, who had millions of dollars spent on his behalf, really racked up the margins among voters, older voters, Baby Boomers in particular. So whether you can draw a straight line between those two things is quite debatable at this moment, but it is something to keep an eye on in these midterm elections. - Seth McLaughlin
So why are people falling for these deepfakes? The images and videos often have a smooth, polished presentation that an aging brain can mistake for authenticity. One of the Florida researchers told Seth it makes them easier for the brain to process, and that ease can be misleading. Layer on what’s called truth bias — the instinct to assume what we see is real — and the risk grows.
Some of it has to do with the fact, you know, they grew up in a different era and you had a few television stations back in the day. And when the well-known newscaster said something, you took it as gold. And so some of those traits carry over to now. - Seth McLaughlin
Truth bias tends to strengthen with age and becomes even more powerful through repetition — the kind of repetition in the form of campaign ads you might see ad nauseam on your TV or phone during an election year.
The political world is still adjusting to this new landscape.
David Martin, who co-founded the ad tech firm AdWave, says the rise of AI has blown apart the high-cost barriers around TV advertising. Previously, it would take between $10,000 to $15,000 to put together an ad. That typically required an ad agency and a production studio, among other details. Martin says the process can now be done for as little as $50, with a finished spot ready to air the next morning.
That shift has opened the door not just for major campaigns, but also for smaller city and county races, including mayors, sheriffs, and other candidates who’d never had a shot at getting on TV before.
Because TV still commands a level of trust that social media never has, older viewers who see a deceptive AI-generated ad on a familiar network may be even less likely to question what they’re watching.
Martin describes where AI is at right now as lawless, frightening, and like the Wild Wild West, where the capabilities of technology are moving faster than the law is, with very little regulation out there.
If this is becoming the Wild West of political advertising, has Congress done anything to rein it in?
A bipartisan bill called the Protect Elections from Deceptive AI Act would ban knowingly using AI-generated audio or video of a federal candidate to trick voters or raise money. But it hasn’t passed. It was referred to its respective committees in the House and Senate last year and has since not been acted on.
Meanwhile, campaigns and candidates aren’t waiting for any clarity or regulations surrounding deepfakes. They’re embracing any chance to alter their or their opponent’s perception among voters.
Former reality TV star Spencer Pratt — you might remember him from the early 2000s series The Hills — has jolted the Los Angeles mayor’s race with a string of AI-generated ads. One of them is a Batman-style spot, casting him as a Caped Crusader trying to save the city from Mayor Karen Bass, who appears movie-style as a Joker-like foil.
And in a contentious GOP Senate primary in Texas, Attorney General Ken Paxton hit Senator John Cornyn with an AI-generated ad showing him dancing with Representative Jasmine Crockett, a liberal Democrat who infuriates the Republican base.
Cornyn and Crockett, the perfect pair.
The trend of trying to deceive older voters isn’t new. Researchers at Princeton and NYU found during the 2016 presidential election, less than 9% of Americans shared links to fake news sites on Facebook, but the behavior was far more common among people 65 and older. That age group shared roughly seven times more false information on Facebook than voters ages 18 to 29.
Still, researchers stress these vulnerabilities among older adults aren’t universal. Older adults who are more analytical or more tech-savvy can spot manipulated content, and a 2020 study by two Harvard researchers found the information fluency of older Americans remains intact, and decades of accumulated knowledge helps them evaluate claims, even if they might forget where they originally learned the information.
Taken together, the collapsing cost of AI ads, the particular vulnerabilities of older voters, and the partisan incentives to provoke outrage have created a perfect environment for this kind of content to spread.
The research suggests older voters may face a double exposure — they might be more likely to be fooled by a convincing deepfake, and also more likely to pass it along once they believe it.
Read more: ‘More real than real’: Why deep-fake political ads are fooling older Americans
Read more political coverage from The Washington Times
Please read our comment policy before commenting.