California Accelerates AI Oversight and Explores Development of an Emergency 'Kill Switch'
California Governor Gavin Newsom issued an executive order on Friday to accelerate the implementation of an independent oversight system for artificial intelligence companies in the state, while simultaneously pushing for the development of an emergency mechanism to halt advanced AI models in the event of serious incidents.
The order directs state agencies to expedite the implementation of two laws passed last week to establish a framework for independent bodies responsible for assessing the safety of AI systems, along with a registry of specialized auditors in this field.
Electricity
Arabs and the peoples of the Middle East
Digital arts and photography
The order also calls for the formation of a group of independent experts to provide recommendations within two months on strengthening AI safety and security laws. These recommendations include subjecting companies developing advanced models to regular audits by independent entities, as well as verifying the safety reports and risk assessments submitted by these companies.
Furthermore, the order explores the development of an emergency "kill switch" for advanced AI models, with its effectiveness subject to continuous verification by an independent body.
According to a statement issued by the Governor of California’s office, the decision follows incidents that raised concerns about the security and safety of AI systems, including an incident linked to Hugging Face, a platform specializing in artificial intelligence and machine learning.
Newsom stated that California is urging the U.S. Congress and the federal government to consider adopting the state’s framework as the basis for national standards to regulate AI. He expressed concern over the lack of effective federal oversight of this technology, particularly given that CEOs of AI companies are calling for its regulation.
He emphasized that the state would not wait to act, but would instead accelerate efforts to establish effective and responsible AI oversight, stressing the need for immediate action.
He added that California is working to strengthen its AI regulatory framework and called for it to be adopted as a baseline standard at the national level.
The decision comes amid debate in the United States regarding the risks of AI and the necessary controls to regulate it, with calls to tighten oversight on advanced models and opposing views against imposing broader restrictions on the development of this technology.
