5 February 2024
WASHINGTON (NEXSTAR) – Some consumer advocates have little faith that Washington will get a handle on deep fakes before Election Day.
Deep fakes are images and videos generated with artificial intelligence and are designed to look real.
Craig Holman with Public Citizen petitioned the Federal Election Commission to ban them in political advertisements last year. The commission is currently reviewing the matter, but he doubts even if they move forward now, they will finalize a new rule before November.
He warns the impact could be detrimental.
“Campaigns can actually depict a candidate doing or saying something that she or he never did, and it looks real. It looks like news, and that will endanger democracy,” Holman said.
The technology has already been used to spread misinformation this election cycle.
During the New Hampshire primary, a fake robocall with a voice impersonating President Joe Biden told voters not to cast their ballots.
The White House says this threat is top of mind and is working with the private sector to find ways to flag deep fake generated content online.
“Watermarking is a great piece of technology with part of that problem, and it’s something that the companies that are driving AI agreed to work on together,” said Science Advisor to President Arati Prabhakar.
Companies like YouTube recently announced policies requiring creators to flag deep fake content.
Federal legislation to outright ban them in elections remains stalled in Congress, despite having bipartisan support.
Senate Majority Leader Chuck Schumer has repeatedly said passing legislation to regulate artificial intelligence is one of the most important priorities Congress must tackle. He has also warned of reacting to quickly and stifling innovation.
Without action from Congress, Holman is now urging states to step up. He says already five states have passed laws regulating deep fakes in elections and several other states have introduced similar bills.