In September, California Governor Gavin Newsom considered 38 AI-related bills, including the highly controversial SB 1047, which the state Legislature sent to him for final approval. He vetoed SB 1047 on Sunday, ending California's controversial AI bill that seeks to prevent AI disasters, but signed more than a dozen other AI bills in September. did. These bills seek to address the most pressing issues in artificial intelligence, from AI risks and deepfake nudes created by AI image generators to Hollywood studios creating AI clones of deceased performers.
“California, home to most of the world's leading AI companies, is leveraging these innovative technologies to help address pressing challenges and the risks they pose,” Governor Newsom's office said in a press release. “We are researching this,” he said.
So far, Governor Newsom has signed 18 AI bills, some of which are the most far-reaching laws on generative AI in the United States to date. Here's what they do:
AI risk
On Sunday, Governor Newsom signed into law SB 896, which requires the California Department of Emergency Services to conduct a risk analysis of the potential threats posed by generative AI. CalOES will work with frontier model companies such as OpenAI and Anthropic to analyze the potential threats of AI to critical state infrastructure and threats that could lead to mass casualty incidents.
training data
Another law Newsom signed this month requires generative AI providers to disclose the data used to train their AI systems in documents published on their websites. AB 2013 will go into effect in 2026, requiring AI providers to specify the source of their datasets, a description of how the data will be used, the number of data points in the set, and whether it contains copyrighted or licensed data. Please, we are obligated to disclose the period etc. Data were collected along with other criteria.
Privacy and AI systems
Newsom also signed AB 1008 on Sunday, clarifying that California's existing privacy laws apply to generative AI systems. This 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 that data for profit. It will be done.
education
This month, Newsom signed AB 2876, which requires the California State Board of Education to consider “AI literacy” in curriculum frameworks and materials for math, science, and history. This means schools in California could begin teaching students not only the basics of how artificial intelligence works, but also the limitations, implications, and ethical considerations of using the technology.
Another new law, SB 1288, requires California's superintendent of schools to establish a working group to study how AI is being used in public school education.
Definition of AI
Newsom signed a bill this month that creates a uniform definition of artificial intelligence in California law. AB 2885 defines artificial intelligence as “artificial intelligence that has varying levels of autonomy and is capable of inferring from received inputs how to produce outputs that affect something physical or virtual, for explicit or implicit purposes. or machine-based systems.” environment. ”
health care
Another bill signed in September is AB 3030, which would require health care providers to disclose when they use generated AI to communicate with patients, especially if those messages include patient clinical information. It's a thing.
Meanwhile, Newsom recently signed SB 1120, which puts limits on how health care providers and health insurance companies can automate services. The law ensures that a qualified physician oversees the use of AI tools in these settings.
AI robocall
Last Friday, Governor Newsom signed a bill requiring robocalls to disclose whether they used an AI-generated voice. AB 2905 aims to prevent new cases of deepfake robocalls similar to Joe Biden's voice that confused many New Hampshire voters earlier this year.
deepfake porn
On Sunday, Newsom signed AB 1831, expanding the scope of existing child pornography laws to include material produced by AI systems.
Last week, Newsom signed two laws addressing the creation and distribution of deepfake nudes. SB 926 would criminalize this practice, making it illegal to blackmail someone with an AI-generated nude image of themselves.
SB 981, also passed Thursday, requires social media platforms to establish channels where users can report deepfake nudes that look like them. The content should then be temporarily blocked while the platform investigates and, if confirmed, permanently removed.
watermark
Also on Thursday, Newsom signed legislation that would allow the public to identify AI-generated content. SB 942 would require widely used generative AI systems to disclose that content is AI-generated in its origin data. For example, every image created by OpenAI's Dall-E will need a small tag in its metadata to indicate that it was generated by AI.
Many AI companies are already doing this, and there are several free tools that allow people to read this provenance data and detect AI-generated content.
election deep fake
Earlier this week, California's governor signed three laws cracking down on AI deepfakes that could influence elections.
One of California's new laws, AB 2655, would require major online platforms like Facebook and X to remove or label election-related AI deepfakes, as well as channels that report such content. It is mandatory to create one. Candidates and elected officials can seek injunctive relief if large online platforms fail to comply with the law.
Another law, AB 2839, targets social media users who post or repost AI deepfakes that could mislead voters about upcoming elections. The law takes effect immediately on Tuesday, and Newsom suggested Elon Musk may be at risk of violating the law.
AI-generated political ads now require full disclosure under California's new law AB 2355. This means President Trump may be forced to post AI deepfakes of Taylor Swift supporting him on Truth Social (she supports 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
Two laws Newsom signed earlier this month, championed by SAG-AFTRA, the nation's largest screen and broadcast actors union, create new standards for California's media industry. AB 2602 requires studios to obtain permission from actors before creating AI-generated replicas of their voices or likenesses.
AB 1836, on the other hand, prohibits studios from creating digital replicas of deceased performers without the consent of their heirs (e.g., legally permitted replicas may not be available in the recent films “Alien'' or “Alien''). (used in Star Wars and even other movies).
SB 1047 vetoed
Governor Newsom still has several AI-related bills to decide on by the end of September. But SB 1047 was not among them, and the bill was vetoed Sunday.
In a letter explaining his decision, Newsom said SB 1047 focuses too narrowly on large-scale AI systems and could “give the public a false sense of security.” California's governor noted that small-scale AI models can be just as dangerous as those covered by SB 1047, and said a more flexible regulatory approach is needed.
During a conversation with Salesforce CEO Marc Benioff earlier this month during the 2024 Dreamforce conference, Newsom may have tipped his hat to SB 1047 and how he thinks about regulating the broader AI industry.
“There is one bill that is kind of substandard in terms of public discussion and awareness, and that is this SB 1047,” Newsom said on stage this month. “What is a demonstrable risk in AI? And what is a hypothetical risk? We can’t solve everything. What can we solve? It’s an approach we use in all areas.”
Check back to this article for the latest information on the AI laws that California governors signed and didn't sign.