WASHINGTON -- Less than three weeks after the Federal Election Commission deadlocked on a request to develop regulations governing so-called deepfake political ads generated using artificial intelligence tools, a non-partisan advocacy group pushing for the new rules is trying again.
Public Citizen will submit a petition to the commission Thursday seeking regulations regarding deliberately misleading campaign communications generated through the use of artificial intelligence.
“AI harms are coming to us from so many different directions, not just in elections, but this might be one of the first visible contact points…where if we don’t have the right regs on the books, we’re going to see immediate consequences,” said Lisa Gilbert, Public Citizen’s executive vice president.
The group’s initial petition was discussed at the FEC’s June 22 meeting. At that time, the three Republican commissioners on the six-person panel voted against the request, saying the petition contain a technical omission. GOP Commissioner Allen Dickerson said the FEC does not have the power to regulate such ads and called on Congress to expand the FEC’s authority.
Public Citizen said its new petition addresses both concerns.
The FEC “without question” has the authority to regulate AI-generated fraudulent ads, Gilbert said. “What we’re talking about is deliberately deceptive content that happens to be created by AI,” she said. “This is something that the FEC can do now, it doesn’t need a statute to force them to act.”
It is unclear when Public Citizen’s petition will be taken up by the FEC.
The use of the technology in ads is more than theoretical, as shown by one way the Republican National Committee responded to President Joe Biden’s announcement in April that he was running for a new term. The RNC released a video, which it disclosed in advance was generated by AI, showing a second Biden term would lead to a dystopian future with cities overrun by crime, a Chinese invasion of Taiwan and bank failures sending markets into free fall.
“Deepfakes’ quality is impressive and already able to fool listeners and viewers,” Public Citizen’s petition states. “Generally, however, on careful examination, it is now possible to identify flaws that show them to be fake. But as the technology continues to improve, it will become increasingly difficult and, perhaps,
nearly impossible for an average person to distinguish deepfake videos and audio clips from
authentic media.”
Legislation requiring political ads created with AI to contain a disclaimer has been introduced in the Senate.
The movement to regulate AI-generated fake campaign ads should not be seen as a partisan issue, Gilbert said. “This is a new tool in elections and there’s no reason to think that Republican consultants or Democratic consultants would use it more or less than the other,” she said. “They will all use it.”
