California signs new AI workplace and consumer protections

California’s announcement combines new worker- and consumer-protection laws with an executive order.[1] The order directs agencies and departments under Newsom’s authority to refer to the technology as “Artificial Intelligence” and “AI,” and it took effect immediately.[2] The workplace measures also address legal work: they prohibit lawyers from handing over core legal tasks, including drafting briefs or providing legal judgment, entirely to AI.[1] Separately, California public higher education institutions must create AI training for students, faculty and staff.[1]

The workplace laws prohibit employers from relying only on AI for disciplinary or termination decisions.[1] They also require employers to provide information if an AI system causes a mass layoff, relocation or termination, and ban workplace surveillance tools in bathrooms.[1] Another reported restriction bars employers from using biometric data to predict a worker’s emotional state.[3] SB 947’s automated-decision-system provisions are scheduled to begin July 1, 2027.[4] Under that law, an employee may request a description of their own data primarily used by the system, and an employer that primarily relied on it for a disciplinary or termination decision must give the affected employee written notice after its use.[4] The law provides for a $500 civil penalty per violation and does not apply to parties covered by a qualifying collective bargaining agreement that explicitly waives its provisions.[4]

The new transparency measures prohibit deleting a digital watermark and allow access to metadata to help identify whether material was AI-generated or user-generated.[1] Another measure requires public comments at local government meetings to come from people rather than AI bots.[1] AB 1609 prohibits large enterprises from presenting customer-service chatbots as human, while exempting certain hospital and healthcare-facility communications related to healthcare services.[5] Together, these provisions address how AI-generated content and automated interactions are identified, as well as whether AI bots may speak in a public-comment setting.[5][1]