California Governor Gavin Newsom issued an executive order directing the Government Operations Agency to investigate amending state law to mandate AI “kill switches” in models developed by California-based firms, verified by an independent organization, sparking sharp criticism from the Free Software Foundation over centralization and user autonomy.
The Executive Order and the Push for Independent Oversight
The executive order, rolled out by Governor Newsom, instructs the California Government Operations Agency to work alongside a panel of experts to study the feasibility of embedding mandatory “kill switches” into artificial intelligence models. This directive follows a string of high-profile security incidents, including the Hugging Face attack, which exposed vulnerabilities in how frontier machine learning models are deployed and maintained.
State actions in California have moved aggressively ahead of federal legislation. Governor Newsom signed Senate Bill 813, establishing a framework for independent verification organizations to assess AI systems for safety and risk. Alongside it, AB 1405 created a state registry for independent AI auditors to evaluate compliance with state law. Newsom’s latest executive order accelerates these implementation timelines and pushes state agencies to explore placing independent verifiers onsite in frontier AI labs.
Software Freedom Concerns Over Centralized Shutdown Mechanisms
While state officials frame these measures as essential consumer and national security safeguards, software freedom advocates argue that government-mandated shutdown mechanisms threaten the foundational principles of computing. The Free Software Foundation raised immediate concerns regarding who ultimately wields administrative control over such a kill switch.
True software freedom requires that users maintain absolute sovereignty over the code running on their systems.
Weighing State-Level Regulation Against Federal Inaction
The regulatory clash highlights a widening gap between California’s aggressive oversight framework and federal inaction. The federal government has enacted no comprehensive legislation requiring AI companies to report dangerous incidents, and the White House has consistently rejected proposals for stringent new federal technology regulations. By establishing SB 53—the Transparency in Frontier Artificial Intelligence Act—alongside SB 813 and AB 1405, California is attempting to build a national blueprint for AI accountability.
As the Government Operations Agency drafts its final recommendations for the governor, software freedom advocates urge policymakers to ensure that individual dignity and user control remain central, rather than cementing centralized corporate and state gatekeeping over modern computing.
