BOSTON – With months to go before the end of the presidential campaign, 2024 is already shaping up to be one of the most turbulent election cycles in history. To complicate matters, deepfake videos are on the rise and voters now have to harness information in more high-tech ways than ever before.

Fake robocall encouraged Democrats not to vote

Last January, thousands of voters received a robocall with a fake version of Biden’s vote generated by artificial intelligence. “Voting this Tuesday only supports Republicans in their quest to elect Donald Trump,” said the call, which urged Democrats do not want to vote in the New Hampshire primaries. “What nonsense,” was the response on the call. Of course, it wasn’t Biden at all.

The twists of the story led the WBZ I-Team to a corner of Manhattan where a street magician named Paul Carpenter was visiting. He said he was the one who created this deep fake message using a computer program. When asked if he regretted it, he hesitated. “Yeah, no,” Carpenter said. “Part of me says I’m upset that I was taken advantage of.”

Although he has no permanent address, Carpenter owns a laptop. In a hotel room, he showed the I-Team how he created the Biden fake. He also shared screenshots he had saved from phone conversations with the political adviser. Steve Kramer, and a Venmo receipt showing that Steve Kramer’s father, Bruce Kramer, paid Carpenter $150. Carpenter said that was the payment he received for creating and delivering the fake Biden message.

Steve Kramer faces a $6 million fine and has been charged with 26 counts of voter suppression, bribery and identity fraud. He worked for the campaign of Dean Phillips, a Democrat who was running against Biden at the time and whose campaign denies having anything to do with the robocall.

Carpenter, who has not been charged, said he never knew the message would be used to manipulate voters. “I would never have participated in that,” he said. “I like America too much.”

Almost 50% of people cannot recognize a fake video

“This is not science fiction. …This is something you can do today,” said David Bau, assistant professor of computer science at Northeastern University. He commissioned student Rohit Gandikota to create deepfake videos for a special WBZ survey.

“The resolution you can work with, the details and the realistic focus have boomed within a year,” said Gandikota.

Because the technology is so new, there’s little research on how many people fall for AI-generated fakes. The I-Team checked with the FBI and the Massachusetts Attorney General’s Office, but they had no records of the number of reported incidents. That’s when WBZ decided to conduct its own investigation. The I-Team interviewed 50 people in five different locations from Boston to Cambridge to Natick. People watched a video in which reporter Christina Hager appeared to say, “We’ve lost several hundred jobs across the MBTA.” In truth, Hager never said those words. It was Massachusetts Governor Maura Healey who actually said them.

The results of the WBZ poll: More than half of the people who saw the video couldn’t spot the fake, but opinions were almost split. While 27 believed it was real, 23 people knew it wasn’t real. The I-Team also conducted a social media poll using an AI-generated voice clip and again found that about half of respondents couldn’t spot what was fake.

Deepfakes could lead to misinformation

And what about the other half? “What we’re really concerned about is that people are getting misinformation and making their voting decision based on that,” said Senator Barry Finegold of Massachusetts. He proposed a bill that would force Massachusetts to join a growing number of states drafting new laws to stop deep fakes. “If they’re using this technology, they have to disclose that and add a disclaimer,” Finegold said.

Back at Northeastern University, Bau urges skepticism about election messages. “It will require some adjustments. We’ve had this problem in other media for some time. People don’t necessarily believe something that’s written on paper.”

CBS News, which owns and operates WBZ-TV, has a dedicated AI department called CBS News Confirmed, which consists of a team of forensic journalists who review questionable videos and reports for veracity and share their findings with viewers.