09-23-2026
Employer Insight: California Governor Gavin Newsom issued a new executive order on September 18, 2026, that directs a group of “world-leading experts” to convene and begin developing a framework for how the state can govern AI.
Newsom told press outlets that California will “speed up” its work on “substantial and responsible AI oversight before it’s too late,” asserting that the stakes are too high for the state to delay action.
The expert committee must give the Governor new guidelines within the next two months to help California “reinforce and strengthen” its AI safety and security laws. Among the proposals the committee will consider are the establishment of independent monitors working inside the top AI companies to verify the safety of AI technology, plans to keep AI safe, and the requirement of a “kill switch” or emergency shutoff mechanism. Some experts are unsure whether “kill switches” would prevent the most serious consequences of AI going rogue.
CA’s executive order comes on the same day that OpenAI announced six additional incidents of “concerning” behavior by its models. In one incident, OpenAI models initiated and hacked onto the Hugging Face platform’s production systems while trying to “cheat” on a cybersecurity test. In another, researchers were able to use Claude AI to hack into OpenAI systems. Earlier in September, an Anthropic researcher publicly resigned because he believed the two major companies were not developing the technology responsibly.
Newsom’s latest executive order builds on earlier orders requiring advanced AI companies to report their safety protocols to the state, and another that explores overhauling labor policies to protect residents from AI job losses.