Last September, California Governor Gavin Newsom reviewed 38 AI-related bills, including the highly controversial SB 1047, which the state Legislature sent to his desk for final approval. He tried to prevent an AI disaster by vetoing SB 1047 on Sunday, but the road is over for California’s controversial AI bill, which has signed more than a dozen other AI bills into law this month. The bill seeks to address some of the most pressing issues in artificial intelligence, from AI risks to deepfake nudes created by AI image generators to Hollywood studios creating AI clones of dead performers.
“California, home to many of the world’s leading AI companies, is committed to leveraging these innovative technologies to solve pressing problems while helping to research risks,” Governor Newsom’s office said in a press release.
To date, Governor Newsom has signed 18 AI bills, some of which are the most expansive laws in the country on generative AI. Here’s what they do:
AI risk
Last Sunday, Governor Newsom signed SB 896 into law, which requires the California Office of Emergency Services to conduct a risk analysis of potential threats posed by generative AI. CalOES will work with leading modeling companies such as OpenAI and Anthropic to analyze potential threats from AI to critical national infrastructure, as well as threats that could lead to mass casualty incidents.
training data
Another law signed by Newsom this month requires generative AI providers to disclose the data used to train AI systems in documents posted on their websites. AB 2013 takes effect in 2026 and requires AI providers to publish the sources of their datasets, a description of how the data will be used, how many data points are in the set, whether it includes copyrighted or licensed data, and for how long. Among other standards, data was collected.
Privacy and AI systems
Newsom also signed AB 1008 on Sunday, specifying that California’s existing privacy laws also apply to generative AI systems. That means that if an AI system like ChatGPT exposes someone’s personal information (name, address, biometric data), California’s existing privacy laws limit how companies can use and profit from that data.
education
Newsom this month signed AB 2876, which requires the California State Board of Education to consider “AI literacy” in math, science and history curriculum frameworks and instructional materials. This means California schools can begin teaching students the basics of how artificial intelligence works, as well as the limits, implications and ethical considerations of using the technology.
Another new law, SB 1288, requires California’s superintendent of schools to create a working group to examine how AI is used in public school education.
AI definition
This month, Newsom signed legislation establishing a uniform definition of artificial intelligence in California law. AB 2885 defines artificial intelligence as “an engineered or machine-based system that has varying levels of autonomy and is capable of inferring from inputs how to produce outputs that can affect something physical or virtual for explicit or implicit goals.” Please specify. environment.”
healthcare
Another bill signed in September is AB 3030, which would require health care providers to disclose if they use generative AI to communicate with patients, especially if those messages include the patient’s clinical information.
Meanwhile, Newsom recently signed SB 1120, which limits how health care providers and health insurers can automate their services. The law ensures that a licensed physician oversees the use of AI tools in these settings.
AI robocall
Last Friday, Governor Newsom signed legislation requiring robocalls to disclose whether they used AI-generated voices. AB 2905 aims to prevent another instance of deepfake robocalls similar to Joe Biden’s voice that confused many New Hampshire voters earlier this year.
deepfake porn
Last Sunday, Newsom signed AB 1831 into law. The bill expands the scope of existing child pornography laws to include material generated by AI systems.
Newsom signed two bills last week that address the creation and spread of deepfake nudes. SB 926 makes it illegal to blackmail someone with an AI-generated nude image of themselves that resembles them, criminalizing the practice.
SB 981, which became law Thursday, requires social media platforms to establish channels for users to report deepfake nudes that look like them. The content should then be temporarily blocked while the platform investigates it, and permanently removed once confirmed.
watermark
Also on Thursday, Newsom signed legislation to help the public identify AI-generated content. SB 942 requires widely used generative AI systems to disclose that their content is AI-generated in the source data. For example, all images generated by OpenAI’s Dall-E now require a small tag in their metadata indicating that they were created by AI.
Many AI companies are already doing this, and there are several free tools to help people read this provenance data and detect AI-generated content.
election deepfake
Earlier this week, California’s governor signed three bills cracking down on AI deepfakes that could influence elections.
One of California’s new laws, AB 2655, would require large online platforms like Facebook and X to remove or label election-related AI deepfakes and create channels to report such content. If large online platforms do not comply with this law, candidates and elected officials can seek injunctive relief.
Another law, AB 2839, takes aim at social media users who post or repost AI deepfakes that could fool voters about the upcoming election. The law took effect immediately Tuesday, and Newsom suggested Elon Musk could be at risk of violating the law.
AI-generated political ads now require full disclosure under California’s new law, AB 2355. That means Trump may not be able to avoid posting an AI deepfake on Truth Social endorsing Taylor Swift (she endorsed Kamala Harris). The FCC has proposed similar disclosure requirements at the national level and has already made robocalls using AI-generated voices illegal.
Actors and AI
The two bills Newsom signed earlier this month, which were being promoted by SAG-AFTRA, the nation’s largest motion picture and broadcast actors’ union, set a new standard for California’s media industry. AB 2602 would require studios to obtain permission from actors before creating AI-generated copies of their voices or likenesses.
Meanwhile, AB 1836 prohibits studios from creating digital replicas of deceased performers without the property’s consent (e.g., legally authorized replicas have been used in the recent “Alien” and “Star Wars” films, as well as other films).
SB 1047 Vetoed
Governor Newsom still has several AI-related bills to decide on before the end of September. But SB 1047 is not one of them. The bill was vetoed on Sunday.
In a letter explaining his decision, Newsom said SB 1047 is too narrowly focused on large-scale AI systems that could “give the public a false sense of security.” The California governor noted that small-scale AI models could be just as risky as those targeted by SB 1047 and said a more flexible regulatory approach is needed.
In a conversation with Salesforce CEO Marc Benioff at the 2024 Dreamforce conference earlier this month, Newsom may have talked about SB 1047 and how he thinks about regulating the AI industry more broadly.
“There is one bill that is rather large in terms of public discourse and awareness. This is SB 1047.” Newsom said on stage this month. “What are provable risks in AI and what are hypothetical risks? I can’t solve everything. What can we solve? That’s the approach we’re taking across the spectrum to this.”
Check back with this article for updates on AI legislation that California’s governor has signed and which AI laws he has not signed.
