California law cracking down on election deepfakes by AI to be tested

SACRAMENTO, CA (AP) — California now has some of the toughest laws in the United States to crack down on election deepfakes ahead of the 2024 election after Gov. Gavin Newsom signed three landmark proposals this week at an artificial intelligence conference in San Francisco.

The state could be among the first to test out such legislation, which bans the use of AI to create false images and videos in political ads close to Election Day.

State lawmakers in more than a dozen states have advanced similar proposals after the emergence of AI began supercharging the threat of election disinformation worldwide, with the new California law being the most sweeping in scope. It targets not only materials that could affect how people vote but also any videos and images that could misrepresent election integrity. The law also covers materials depicting election workers and voting machines, not just political candidates.

Among the three law signed by Newsom on Tuesday, only one takes effect immediately to prevent deepfakes surrounding the 2024 election. It makes it illegal to create and publish false materials related to elections 120 days before Election Day and 60 days thereafter. It also allows courts to stop the distribution of the materials, and violators could face civil penalties. The law exempts parody and satire.

The goal, Newsom and lawmakers said, is to prevent the erosion of public trust in U.S. elections amid a “fraught political climate.”

The legislation is already drawing fierce criticism from free speech advocates and social media platform operators.

Elon Musk, owner of the social media platform X, called the new California law unconstitutional and an infringement on the First Amendment.

Hours after they were signed into law, Musk on Tuesday night elevated a post on X sharing an AI-generated video featuring altered audios of Vice President and Democratic presidential nominee Kamala Harris. His post of another deepfake featuring Harris prompted Newsom to vow to pass legislation cracking down on the practice in July.

“The governor of California just made this parody video illegal in violation of the Constitution of the United States. Would be a shame if it went viral,” Musk wrote of the AI-generated video, which has the caption identifying the video as a parody.

But it’s not clear how effective these laws are in stopping election deepfakes, said Ilana Beller of Public Citizen, a nonprofit consumer advocacy organization. The group tracks state legislation related to election deepfakes.

None of the law has been tested in a courtroom, Beller said.

The law’s effectiveness could be blunted by the slowness of the courts against a technology that can produce fake images for political ads and disseminate them at warp speed.

It could take several days for a court to order injunctive relief to stop the distribution of the content, and by then, damages to a candidate or to an election could have been already done, Beller said.

“In an ideal world, we’d be able to take the content down the second it goes up,” she said. “Because the sooner you can take down the content, the less people see it, the less people proliferate it through reposts and the like, and the quicker you’re able to dispel it.”

Still, having such a law on the books could serve as a deterrent for potential violations, she said.

Newsom’s office didn’t immediately respond to questions about whether Musk’s post violated the new state law.

Assemblymember Gail Pellerin, author of the law, wasn’t immediately available Wednesday to comment.

Newsom on Tuesday also signed two other laws, built upon some of the first-in-the-nation legislation targeting election deepfakes enacted in California in 2019, to require campaigns to start disclosing AI-generated materials and mandate online platforms, like X, to remove the deceptive material. Those laws will take effect next year, after the 2024 election.