Experts warn of spread of AI and ‘deepfake’ images ahead of elections

NASHVILLE, Tenn. (WKRN) — The old saying goes you can’t trust everything you read on the internet, but it’s becoming increasingly clear you can’t trust everything you see either.

Artificial intelligence and so-called “deepfake” videos have proliferated across social media platforms. With the election approaching, demands for regulation have increased.

Dozens of shots fired at the home of the pastor of the Global Vision Bible Church

With one click online, you can see a post that is meant to touch your heart. It may show a picture of a 3-year-old boy. The woman who wrote it may be asking for help. But in reality, the entire post is fake.

Lisa Fazio, an assistant professor of psychology and human development at Vanderbilt University, said these types of posts are meant to spread.

“People tend to create (deepfake posts) that people want to share and spread,” said Lisa Fazio, assistant professor of psychology and human development. “Information about a missing child — you want to make sure the child is found.”

Fazio explained that the possibilities of artificial intelligence can seem daunting, especially since the technology has the ability to manipulate images without being noticed.

That’s why some police departments across Middle Tennessee have issued the same warning as the Mt. Juliet Police Department, urging people to be on the lookout for fake posts.

“Humans have a hard time spotting misinformation in the wild,” Fazio added. “It’s hard for us to notice when something contradicts our prior knowledge or what we’re supposed to think about the world, and images are a really powerful form of information. Deepfakes, edited images, or images just used out of context can be a really powerful way to change our beliefs.”

But it’s not just social media that’s happening. AI has now entered politics. With the presidential election in November, some are now concerned about the lack of federal regulations on deceptive deepfakes in political ads.

For months, Public Citizen has been pushing the federal government to ban fake political images, videos and voice recordings, which are becoming increasingly difficult to detect.

Nolensville city officials want to make changes to one road after a teenager was hit by a car

“It’s entirely possible that we’ll see late-night deepfakes that show the candidate drunk, saying something racist, or behaving in an outrageous manner, even though he never said any of those things,” explained Robert Weissman, co-chair of Public Citizen.

The Federal Communications Commission is working to finalize rules requiring all television and radio ads created using artificial intelligence to include a disclaimer.

For the latest news, weather, sports events and video coverage, visit WKRN News 2.