California passes package of laws to combat election deepfakes

California Governor Gavin Newsom passed a tough new law to crack down on politically-themed artificial intelligence deepfakes during elections. 

It comes only weeks after Elon Musk re-posted a parody of a Kamala Harris campaign ad on X that garnered millions of views and used AI-powered voice manipulation to make it seem Harris called herself an incompetent presidential candidate. 

In late July, Newsom specifically pointed to Musk’s post and vowed to sign a bill “in a matter of weeks” banning the practice.

Kamala Harris Campaign Ad PARODY pic.twitter.com/5lBxvyTZ3o

— Mr Reagan 🇺🇸 (@MrReaganUSA) July 26, 2024

The new law, known as AB 2839 , takes effect immediately and effectively bans people and groups from knowingly sharing election deepfakes and other “materially deceptive content.” 

Existing laws only prohibit the distribution of materially deceptive audio or visual media of a candidate within 60 days of an election. The new law expands the ban to 120 days before an election in California and, in some cases, 60 days after an election is held.

Newsom also signed two other laws aimed at taking down political AI deepfakes on the same day, though they won't be in effect until January.

Source: Gavin Newsom

One of the laws, AB 2355 , will require labels to be put on political advertisements generated or substantially altered with AI. 

The other, AB 2655 , requires social media platforms with more than one million California users, such as Facebook and X, to block deceptive content relating to elections during specified periods. 

It will also make platforms liable for failing to remove content within 72 hours of receiving a report from a user.

All three laws only apply to California, and it is understood that the governor signed the bills during a fireside chat with Salesforce CEO Marc Benioff at the Dreamforce conference in San Francisco. 

“It’s very different from other bills that have been put forth,” Ilana Beller, an organizing manager of the democracy team at consumer advocacy organization Public Citizen told The New York Times. 

Related: AI deepfake attacks will extend beyond videos and audio — Security firms

Meanwhile, Newsom signed an additional two bills, AB 1836 and AB 2602, on Sept. 17, which aim to give greater protection to performers over the use of their digital likenesses. 

Gavin Newsom (bottom left) signs AI bills to protect performers against AI misuse. Source: Instagram

The fake Harris campaign parody ad, posted on July 26, has reached 25.4 million views on X alone and is still available on the platform today.  

Musk later defended his retweeting of the original post at the time, mocking Newsom:

“I checked with renowned world authority, Professor Suggon Deeznutz, and he said parody is legal in America”

Magazine: AI drone ‘hellscape’ plan for Taiwan, LLMs too dumb to destroy humanity: AI Eye