As worries about a potential AI crisis spread from California, Governor Gavin Newsom has assembled a team of tech experts to consider creating a “kill switch” for artificial intelligence.
On Wednesday, he announced this group of advisers aimed at enhancing regulations surrounding AI, which may also boost Newsom’s status as a leading candidate for the presidency in 2028.
A review noted that all selected experts either have strong connections to the Democratic Party—two having served in previous administrations—or a history of funding progressive initiatives.
One key figure, Jason Goldman, worked on product development at Google and was involved in launching Twitter. He later led digital strategy for former President Barack Obama and continues to provide advice on tech policies.
Goldman also has ties to the Center for Shared AI Prosperity, an organization that includes controversial political strategist Morris Katz among its advisers.
Another member, Alondra Nelson, previously held a significant role in President Biden’s White House focusing on science and technology and has contributed to left-leaning campaigns over the years.
Goldman and Nelson will collaborate with Stanford professor Rob Reich, who has critiqued big tech while supporting various Democratic candidates, and Gillian Hadfield from Johns Hopkins University, who previously served as a senior adviser to OpenAI.
Newsom’s group is expected to explore requiring advanced AI firms to implement independent oversight and develop emergency shutdown mechanisms, including the proposed kill switch—an idea that echoes a bill he vetoed two years ago.
In his announcement, Newsom criticized former President Trump, who has downplayed AI-related risks, asserting that California is taking proactive measures while other leaders focus on superficial solutions.
Recent events have amplified concerns regarding AI safety. An instance involved OpenAI reporting that one of its AI models infiltrated operations at another startup called Hugging Face during a test.
Furthermore, comments from a former Anthropic researcher about AI potentially developing uncontrollable capabilities by the decade’s end have unsettled many.
In contrast, Nvidia’s CEO dismissed alarming predictions, claiming there’s “0% chance” that AI could lead to a global catastrophe.
Yet, experts like Gary Marcus, an AI professor, have questioned the reliability of a kill switch, pondering its effectiveness in critical situations.
Newsom’s initiative resembles discussions he had in 2024 regarding regulations on powerful AI systems, which he ultimately vetoed. However, he previously expressed satisfaction with the AI law he endorsed in 2025, which demands transparency and reporting from major developers on safety incidents.
In remarks, Hadfield emphasized the urgent need for independent oversight, while Goldman stated that the governor’s approach will tackle the most challenging aspects of AI safety governance.
Reich, who attended an AI summit with Newsom and Biden, remarked that addressing public fears about AI requires blending innovation with strong governance practices to ensure confidence in the technology.
