Here’s What’s Illegal Under California’s 8th (and Growing) New AI Law

California Gov. Gavin Newsom is currently considering 38 AI-related bills, including the highly controversial SB 1047, which the state legislature has sent to his desk for final approval. These bills attempt to deal with essentially the most pressing issues in AI: from futuristic AI systems posing existential risks, deepfake nudes from AI image generators, to Hollywood studios creating AI clones of dead performers.

“California, home to many of the world’s leading AI companies, is working to leverage these breakthrough technologies to help solve pressing problems while also investigating the risks they pose,” Newsom’s office said in an announcement. press release.

Governor Newsom has signed eight of them into law to this point, a few of that are essentially the most far-reaching AI laws in American history.

Deepfake nudity

On Thursday, Newsom signed two bills into law addressing the creation and distribution of pretend nude images. SB926 criminalizes this act by making it illegal to blackmail someone with AI-generated nude photos that resemble that person.

SB981which also went into effect Thursday, requires social media platforms to establish channels for users to report deepfake nudes that resemble them. The content should be temporarily blocked while the platform investigates it, and if confirmed, it should be permanently removed.

Watermarks

Also on Thursday, Newsom signed a bill that may help the general public discover content generated by artificial intelligence. SB942 requires commonly used generative AI systems to disclose that they’re AI-generated of their content’s provenance data. For example, all images created by OpenAI’s Dall-E now need a small tag of their metadata saying they’re AI-generated.

Many AI corporations are already doing this, and there are a lot of free tools available that might help people decode provenance data and detect AI-generated content.

Election deepfakes

Earlier this week, California’s governor signed three bills geared toward combating artificial intelligence (AI)-based misinformation that would influence elections.

One of the brand new laws in California, AB2655requires major web platforms like Facebook and X to remove or label election-related AI deepfakes, in addition to create channels to report such content. Candidates and elected officials can seek a court order if a serious web platform fails to comply with the bill.

Another law, AB2839targets social media users who post or repost AI deepfakes that would deceive voters in regards to the upcoming election. The law went into effect immediately on Tuesday, and Newsom suggested that Elon Musk could possibly be in danger for violating it.

AI-generated political ads now require public disclosure under latest California law, AB2355. That means Trump may not have the opportunity to get away with posting AI deepfakes of Taylor Swift endorsing him on Truth Social in the long run (she endorsed Kamala Harris). The FCC has proposed an analogous disclosure requirement nationally and has already illegalized robocalls using AI-generated voices.

Actors and AI

The two laws signed Tuesday by Newsom, SAG-AFTRA, the biggest film and tv actors union within the country, establish latest standards for California’s media industry. AB2602 requires film studios to acquire an actor’s consent before creating an AI-generated replica of their voice or likeness.

Meanwhile, 1836-01-01 prohibits studios from creating digital replicas of deceased artists without the consent of their heirs (e.g., legally approved replicas were utilized in the recent Aliens and Star Wars movies, in addition to other movies).

What’s left?

Governor Newsom has 30 more AI-related bills to act on before the top of September. During a conversation with Salesforce CEO Marc Benioff on Tuesday during Dreamforce 2024, Newsom was capable of tip his hat to SB 1047 and the way he thinks about regulating the AI ​​industry more broadly.

“There’s one bill that’s a little bit over the top in terms of public discourse and awareness; it’s SB 1047,” Newsom said on stage Tuesday. “What are the proven risks in AI and what are the hypothetical risks? I can’t solve everything. What can we solve? And that’s why we’re taking that approach across the board.”

Check out this text to search out out what AI bills the California governor will and won’t sign.