MARYLAND — They say don't believe everything you see on the internet. That phrase has never been more true, but it's also never been more difficult to determine what's real and what's fake. And as artificial intelligence becomes more intelligent -
"I think it's gonna get worse and worse - in terms of, the technology is gonna make it harder for us to visually check if something is true or not. On the flip side, we're gonna be increasingly trained into maybe not sharing those things," University of Maryland professor of media and democracy Daniel Trielli told WMAR-2 News in a recent interview.
But right now, deep-fakes and AI-generated images are shared wildly. And this election season, bad actors are seizing on the opportunity to either influence voters, or simply to generate clicks and money.
Some images are intentionally designed to deceive people. Other images are simply meant to evoke a certain feeling or association - like this photo showing Donald Trump running away from police as they attempt to arrest him, or this one showing Kamala Harris speaking in front of communist flags.
Making pictures of Trump getting arrested while waiting for Trump's arrest. pic.twitter.com/4D2QQfUpLZ
— Eliot Higgins (@EliotHiggins) March 20, 2023
— Donald J. Trump (@realDonaldTrump) August 18, 2024
"For most people, you can see that image and you can see that it's not truthful, that it's something that didn't happen. But the idea there was to embed that feeling of - she's a communist - into her image," Trielli explained.
Trielli says, often the content is tailored to tap into people's already-held beliefs, or weaponize their mistrust of government.
"Everyone is more likely to believe something that confirms their prior belief or their prior philosophy than something that confronts that. And people who deal with misinformation or disinformation are very aware of that."
It's hard to tell where the content is coming from. Anyone with access to the Internet can create a fake image using AI.
"It's possible that there are some specific content farms either in the U.S. or abroad that are pumping in disinformation throughout the country and in fact throughout the world to try and impact elections."
The Office of the Director of National Intelligence told ABC News in September that Russia and Iran were using AI "convincingly tailor synthetic content" to influence the election.
There are some tell-tale signs you can look out for when trying to figure out if an image is AI-generated.
"Generative AI can't do hands very well, so the fingers might be wonky or something like that," Triell said. An example of that can be seen in this photo of Trump praying in a church with six fingers on one of his hands.
On Truth Social yesterday, Donald Trump shared a picture of himself praying with six fingers -- a clear indication the image is AI-generated. pic.twitter.com/ZfA8ygkXem
— Bill McCarthy (@billdmccarthy) January 23, 2024
"There's usually like a glossy, shiny style to generative AI. But the reality is that those signs are disappearing as technology is evolving," Trielli warns.
That's means it's on us - to think before we share. Trielli says most people don't realize how much power they wield online.
"We're all publishers and we all should have some degree of responsibility for what we're sharing."
Trielli expects to see a lot of fake images on Election Day in order to confuse people about where and how to vote, in addition to content about voter fraud in the days leading up to the election and the days after.
The Maryland State Board of Elections has a rumor control page on its website, where you can find official information, and also report disinformation.