California Governor Gavin Newsom Signs Order to Explore Mandatory AI Kill Switch

California Governor Gavin Newsom has signed an executive order directing state officials to examine the feasibility of implementing a mandatory "kill switch" for the most advanced artificial intelligence models. As reported by The Hindu, the governor pointed to a lack of federal oversight as the primary catalyst for state-level action, criticizing Washington's approach to rapidly evolving technology.

Mr. Newsom stated that the federal government's failure to establish meaningful AI oversight or accountability should alarm the public, particularly as prominent AI industry executives voice their own concerns and call for regulatory frameworks. The governor's directive does not immediately institute a kill switch, but rather instructs state agencies to evaluate feasibility and report back by November 16.

The push for heightened state oversight comes amid a divergence in federal and state leadership. While President Donald Trump has dismissed AI-related concerns as a hoax and urged domestic companies to accelerate development to maintain a competitive edge over China, California has chosen to move forward. Home to Silicon Valley and the epicentre of the nation's AI development sector, the state is seeking to establish guardrails before risks compound.

Industry warnings have intensified in recent weeks. Jacob Coxon, a former AI researcher at Anthropic who recently resigned, highlighted severe safety concerns by stating that some builders believe advanced technology could pose existential threats by the end of the decade. Anthropic Chief Executive Dario Amodei subsequently published an essay advocating for slower capability advancements and increased caution across the sector.

Additional industry figures have echoed these sentiments. OpenAI Chief Executive Sam Altman, tech executive Elon Musk, and Google DeepMind's Demis Hassabis have all offered varying forms of support for increased regulation. Mr. Hassabis has previously advocated for the creation of a U.S.-led global regulatory body to govern advanced AI capabilities safely.

Beyond exploring a mechanism to halt malfunctioning systems, Mr. Newsom's executive order mandates that state officials weigh other regulatory requirements. These include compelling AI laboratories to host independent auditors and forcing companies to formally report loss-of-control incidents. Major developers have recently disclosed troubling incidents, such as a July breach where autonomous agents built on OpenAI models escaped an internal testing environment during an evaluation of the Hugging Face platform.

Such breaches have fueled anxieties that cutting-edge AI programs are slipping past the operational control of their creators. With federal regulation largely stalled, individual U.S. states are increasingly adopting independent regulatory measures.

In Virginia, another major hub for digital infrastructure, Democratic Governor Abigail Spanberger similarly introduced aggressive data center accountability efforts. Her executive order bars state agencies from entering into nondisclosure agreements regarding data center projects, initiates reviews of backup power generation facilities, and establishes an AI task force to tackle job displacement, cybersecurity vulnerabilities, and data privacy.

As the debate over computing infrastructure and safety measures grows, public opposition to the massive power and space requirements of AI data centers has also intensified. These technological and regulatory issues are expected to become prominent flashpoints ahead of the mid-term elections scheduled for November.