Experts warn AI political misinformation surging, impact on elections

 

MADISON, Wis. (CBS 58) -- Almost two months out until Election Day, experts are warning about a flood of misinformation on social media as millions of images and videos can easily slip through the cracks despite advancements in technology.

Part of that has to do with artificial intelligence and the ability to create "deep fakes," as the tools can make it much easier and cheaper to create content that could mislead voters.

Democrats and Republicans in Congress are trying to take steps to regulate AI, but they've been unable to approve any legislation.

It's left some states, including Wisconsin, to put their own in place by implementing laws requiring campaigns to disclose when AI is used in political ads.

Misinformation experts say while it's a step in the right direction, that alone won't solve the problem.

"Whatever we like best [on social media], we're going to get more and more of it. And if that's misinformation, we're going to get targeted by it more," said Dietram A. Scheufele, an expert of misinformation and professor of life sciences communication at the University of Wisconsin-Madison.

"That's the pernicious dynamic that's going to get worse and worse."

Scheufele and others are bracing for an influx of AI political misinformation they say begins to surge just days leading up to the November election.

Over the last month, fake images of women wearing "Swifties for Trump", and a picture of Vice President Kamala Harris depicted as a communist leader posted on "X" by former President Donald Trump, have gone viral.

Other mechanisms like robocalls have also grown in popularity. In January, thousands of New Hampshire voters received deceptive calls using artificial intelligence to mimic President Joe Biden's voice to spread misinformation that would prevent them from voting in the state's presidential primary.

While there are some bad actors using AI, experts say it can also be a tool to combat false information by analyzing patterns, language and context.

Oren Etzioni, an artificial intelligence expert and professor emeritus at University of Washington, created a nonprofit startup that works to depict truth from fiction.

True Media joins a growing list of websites helping people spot AI-generated deep fakes.

"When bad actors use AI for nefarious means, one of the best lines of defense is using AI to thwart their attempts," said Etzioni. "True Media is using good AI to fight bad AI."

Ahead of the election, hacking operations also have U.S. officials on edge.

Earlier this month, the FBI and U.S. intelligence agencies revealed the Iranian government was behind the hack and lead operation targeting the Trump campaign and the attempted attack against the Biden-Harris campaign.

Experts say while officials are much more prepared this time around than 2016, when Russia hacked Hillary Clinton's emails and leaked them to the press, intelligence officers are also walking a fine line of trying to be open about these incidents without creating panic about election interference.

Scheufele said because foreign actors are "very hard to control," the U.S. "will always be playing catchup" because they are often one step ahead.