Voters are already having trouble separating fact from fiction, so the arrival on the scene of Artificial Intelligence (AI) in the form of deepfakes is deeply disturbing to those worried about democracy. 

New York state Assemblyman Clyde Vanel, who serves as the chair of the Subcommittee on Internet and New Technology, created the following post on X, formerly known as Twitter, as an example of why AI needs to be regulated.

“I really believe that 2024 is going to be the first AI election we’ve had in the United States,” said Lawrence Norden, senior director of the Elections and Government Program at the Brennan Center for Justice. “AI is very good at imitating authoritative voices.”

Norden is concerned that the 2024 election cycle will include deepfakes of candidates as well as attempts at voter suppression.

“In the past, there have been robocalls, but you can imagine synthetic audio imitating election officials and others, looking to confuse voters about where to vote (or) when to vote,” he said.

Other issues the Brennan Center is on the lookout for include fake communications between the public and candidates, and election security issues where there might be attempts to break into election systems. 

Lawmakers in at least five states — California, Michigan, Minnesota, Texas, and Washington — have passed new laws attempting to criminalize or impose civil penalties on candidates, political action committees and others who create ads or other types of communications that imitate candidates in a way that is false. 

Assemblyman Clyde Vanel has introduced at least two bills dealing with AI:

• A6790 prohibits the creation and dissemination of synthetic media within sixty days of an election with intent to influence the outcome of the election. 
• A8195 enacts the “advanced artificial intelligence licensing act”, which provides for the regulation and licensing of high-risk advanced AI systems. 

“We are really concerned about high-risk AI situations, including elections, election infrastructure, health care and weapons systems,” Vanel told Capital Tonight.

“Perhaps it’s not surprising that the thing both in Congress and in state Legislatures across the country that legislators seem most interested in are deepfakes that can imitate them in political communications,” Norden said.

Currently, there are eight bills in Congress to attempt to regulate deepfakes.

Other countries, including Slovakia, have seen deepfakes turn up just days before an election when there wasn’t enough time to debunk the lies generated.

“There’s audio of a candidate, right before the election, the weekend before, saying that he was going to raise taxes in, what was supposedly a private conversation. It was all false. And it went viral,” said Norden. “I think we can pass legislation, but most importantly is that we educate the public ahead of time.”