New California Laws Could Kill Election Deepfakes on Your Social Media Feeds

PCMag editors select and review products independently. If you buy through affiliate links, we may earn commissions, which help support our testing.

California Governor Gavin Newsom signed three bills this week aimed at limiting the spread of election-related deepfakes on social media.

“Safeguarding the integrity of elections is essential to democracy, and it’s critical that we ensure AI is not deployed to undermine the public’s trust through disinformation – especially in today’s fraught political climate," says Newsom.

Bill 1: You Can't Share Deceptive, AI-Generated Posts Near an Election

Only one, AB 2839, takes effect before the 2024 presidential election. The "urgency measure" expands an existing law that prevents people or groups from "knowingly distributing" deceptive, AI-generated election materials. It also allows election officials, candidates, and others to sue to prevent the distribution of such material. The existing law prevents distributing materials within 60 days of an election. The new provision prohibits it within 120 days of an election in California and, in certain cases, 60 days after an election.

Bill 2: Political Campaigns Must Disclose AI Advertisements

The other two measures go into effect in January 2025. The first, AB 2355, requires candidates to disclose when their campaign ads use AI-generated, or "substantially altered," content.

Donald Trump has favored AI-generated images on his social media feeds. Four days after X unveiled an AI image generator that readily creates outrageous images of political candidates, Donald Trump posted an AI-generated image of Vice President Kamala Harris on stage at the Democratic National Convention under a hammer and sickle flag with the word "Chicago" lit up in red over the crowd. The post did not include an AI-generated disclosure.

California is the first state to include AI in its campaign transparency rules, says Assemblymember Wendy Carrillo, who sponsored the bill.

"As these technologies become more accessible and are used in political campaigns, their impact on democracy requires urgent action," says Carrillo. "Free speech and political expression are a cornerstone of our democracy, but we cannot lose sight of our humanity amid the advancement of artificial intelligence."

Bill 3: Social Media Companies Must Label, Remove AI Deepfakes

A third law, AB 2655, which also takes effect in January, requires large social media platforms to label or remove AI deepfakes within 72 hours after receiving a complaint. The site must create mechanisms to receive those complaints. If the site does not address the complaints, then candidates, election officials, and government officials can require it to do so with legal action.

The Defending Democracy from Deepfake Deception Act of 2024 does not apply to satire or parody posts, and only applies to specified periods of time, such as leading up to an election.

AB 2655 could help standardize content moderation across social platforms, rather than leaving it to each site to decide what stays and goes. In February, a fake video of President Joe Biden spread on Facebook, which Meta's oversight board decided to keep up. Another one of VP Harris made the rounds on X in July, thanks in part to CEO Elon Musk retweeting it to his 198 million followers.

“AI-generated deepfakes pose a clear and present risk to our elections and our democracy," says Assemblymember Marc Berman, who sponsored the bill. "AB 2655 is a first-in-the-nation solution to this growing threat. The new law is a win for California’s voters, and for our democracy."

California seeks to set an example for the nation's AI regulations with these laws, but it's not alone. Twenty-four states have either already passed election-related deepfake laws, or are awaiting signature on them, according to The New York Times, citing data from Public Citizen.

Hollywood Gets Its AI Bill, Elon Musk May Not

Also this week, Newsom signed two bills protecting Hollywood actors from having their digital likeness reproduced by AI-generated audio and visual productions, including performers who are deceased.

“It is a momentous day for SAG-AFTRA members and everyone else because the AI protections we fought so hard for last year are now expanded upon by California law thanks to the legislature and Governor Gavin Newsom," says SAG-AFTRA President Fran Drescher. "They say as California goes, so goes the nation!”

But one final, controversial bill remains on Newsom's desk, which has generated debate among US lawmakers and big tech companies alike.

would introduce more safety and transparency requirements for large AI systems like ChatGPT. It has the support of Elon Musk and Anthropic CEO Dario Amodei, who say it's needed to hold large systems accountable and prevent an AI catastrophe. Others, such as Nancy Pelosi, say it will stifle competition.

The bill passed a final state assembly vote and awaits Newsom's veto or approval by Sept. 31. This week, he expressed concerns about the "chilling effect" the bill could have on the industry, as first reported by Bloomberg.