Skip to main content

California governor vetoes expansive AI safety bill

California Governor Gavin Newsom speaking at a lecturn.
Gage Skidmore / Flickr

California Gov. Gavin Newsom has vetoed SB 1047, the Safe and Secure Innovation for Frontier Artificial Models Act, arguing in a letter to lawmakers that it “establishes a regulatory framework that could give the public a false sense of security about controlling this fast-moving technology.”

“I do not believe this is the best approach to protecting the public from real threats posed by the technology,” he wrote. SB 1047 would have required “that a developer, before beginning to initially train a covered model … comply with various requirements, including implementing the capability to promptly enact a full shutdown … and implement a written and separate safety and security protocol.”

Recommended Videos

However, Newsom noted that 32 of the top 50 AI companies are based in California, and that the bill would focus on only the largest firms. “Smaller, specialized models may emerge as equally or even more dangerous than the models targeted by SB 1047,” he stated.

“While well-intentioned, SB 1047 does not take into account whether an Al system is deployed in high-risk environments, involves critical decision-making or the use of sensitive data,” Newsom wrote. “Instead, the bill applies stringent standards to even the most basic functions – so long as a large system deploys it.”

SB 1047 sparked heated debate within the AI industry as it made its way through the legislature. OpenAI stridently opposed the measure, resulting in researchers William Saunders and Daniel Kokotajlo publicly resigning in protest, while xAI CEO Elon Musk came out in favor of the bill. Many in Hollywood also expressed support for SB 1047, including J.J. Abrams, Jane Fonda, Pedro Pascal, Shonda Rhimes, and Mark Hamill.

“We cannot afford to wait for a major catastrophe to occur before taking action to protect the public. California will not abandon its responsibility. Safety protocols must be adopted. Proactive guardrails should be implemented, and severe consequences for bad actors must be clear and enforceable,” Newsom wrote. However, “ultimately, any framework for
effectively regulating Al needs to keep pace with the technology itself.”

Monday’s announcement comes less than a month after the governor signed AB 2602 and AB 1836, both backed by the SAG-AFTRA union. AB 2602 requires performers grant informed consent prior to using their “digital replicas,” while AB 1836 strengthened protections against ripping off the voice and likeness of deceased performers.

Andrew Tarantola
Andrew Tarantola is a journalist with more than a decade reporting on emerging technologies ranging from robotics and machine…
OpenAI’s Sora was leaked in protest over allegations of ‘art washing’
An AI image portraying two mammoths that walk through snow, with mountains and a forest in the background.

OpenAI's unreleased Sora video generation model was leaked Tuesday by a group protesting the company's "art washing" actions, per a post from X user @legit_rumors.

The group, calling themselves Sora PR Puppets, reportedly had gained early access to the Sora API. Through that, they leveraged authentication tokens to create a front-end interface enabling anyone to generate video clips with the model. While the project only remained online for around three hours before Hugging Face (or possibly OpenAI itself) revoked access, several users managed to publish their creations to social media sites.

Read more
Is AI already plateauing? New reporting suggests GPT-5 may be in trouble
A person sits in front of a laptop. On the laptop screen is the home page for OpenAI's ChatGPT artificial intelligence chatbot.

OpenAI's next-generation Orion model of ChatGPT, which is both rumored and denied to be arriving by the end of the year, may not be all it's been hyped to be once it arrives, according to a new report from The Information.

Citing anonymous OpenAI employees, the report claims the Orion model has shown a "far smaller" improvement over its GPT-4 predecessor than GPT-4 showed over GPT-3. Those sources also note that Orion "isn’t reliably better than its predecessor [GPT-4] in handling certain tasks," specifically coding applications, though the new model is notably stronger at general language capabilities, such as summarizing documents or generating emails.

Read more
OpenAI’s robotics plans aim to ‘bring AI into the physical world’
The Figure 02 robot looking at its own hand

OpenAI continued to accelerate its hardware and embodied AI ambitions on Tuesday, with the announcement that Caitlin Kalinowski, the now-former head of hardware at Oculus VR, will lead its robotics and consumer hardware team.

"OpenAI and ChatGPT have already changed the world, improving how people get and interact with information and delivering meaningful benefits around the globe," Kalinowski wrote on a LinkedIn announcement. "AI is the most exciting engineering frontier in tech right now, and I could not be more excited to be part of this team."

Read more