Illinois and California Governors Order New AI Safety Cabinets

Governors in Illinois and California have moved to impose stricter controls on artificial intelligence, citing urgent risks from unchecked model development. Within four days, both state leaders signed executive orders aimed at creating dedicated oversight structures and accelerating safety requirements for major technology companies.

The actions follow recent concerns raised about the capabilities of frontier models from leading AI developers. By taking state-level action, the governors are bypassing potential federal delays to implement immediate safeguards for public infrastructure and data.

Illinois creates dedicated AI safety cabinet

Illinois Governor JB Pritzker established the Illinois Artificial Intelligence Cabinet to centralize expertise on AI risks. This body will advise state government on incident response, infrastructure protection, and accountability measures. It brings together leaders from agencies such as the Department of Innovation and Technology, the Emergency Management Agency, and the State Police.

The cabinet will also include external experts from academia, law, and ethics. Pritzker stated that while AI has transformative potential, the state must listen to experts urging caution. He emphasized that Illinois will not wait for federal action to protect residents, asserting that state responsibility requires immediate preparation and protection.

California accelerates independent verification timelines

California Governor Gavin Newsom signed an order to speed up the implementation of independent oversight for AI companies. The executive order shortens the deadline for establishing regulations for Independent Verification Organizations, or IVOs, from January 2028 to May 2027. These third-party firms will audit frontier AI labs to ensure safety measures are actually in place.

The order also tasks state agencies with drafting potential legal amendments within eight weeks. These proposals include requiring IVOs to be embedded in developer labs for periodic audits and verifying the accuracy of safety reports filed by companies. The goal is to create a verifiable framework that prevents unmonitored expansion of powerful models.

Kill switch proposals face implementation hurdles

A central component of both states' strategies is the development of an emergency shutoff mechanism, often called a kill switch. This feature would allow for the immediate cessation of a frontier model's operations if it poses a danger. However, the technical feasibility and legal enforceability of such a switch remain open questions that regulators must resolve.

The Transparency Coalition reported that these executive orders represent a significant shift toward state-led regulation. By mandating third-party verification and onsite audits, the governors are attempting to close gaps that voluntary industry standards have left open. The trade-off is increased regulatory burden on developers, which may slow innovation while enhancing safety.