Newsom signs law barring AI from independently making licensed medical decisions in California
Governor Gavin Newsom has signed Assembly Bill 1979, which bars hospitals and medical offices in California from letting artificial intelligence independently perform medical tasks that require a professional license – such as prescribing medications or ordering tests. The law takes effect January 1, 2027, and lands in the middle of an ongoing dispute over algorithmic care tools at Kaiser Permanente. But according to the Mercury News, it remains unclear which AI tools, if any, the law actually prohibits.
What the law says
Assembly Bill 1979 was signed by Governor Gavin Newsom last week, relative to the Mercury News' October 10 publication, and takes effect January 1, 2027. According to the Mercury News' reporting, it prohibits hospitals and medical offices from using artificial intelligence to independently perform medical tasks that require a professional license, with prescribing medications and ordering tests as named examples (Mercury News).
The law also contains a second requirement: healthcare providers must take "reasonable steps" to ensure that licensed professionals can exercise independent judgment when AI tools help guide patient care.
The unions push back against the trend
The bill was carried by Assemblymember Mia Bonta (D-Oakland) and was sponsored by the California Nurses Association. "AI is rapidly integrating into our health care system and reshaping our personal experience with it," Bonta said in a statement cited by the Mercury News.
For the union, the issue is what its members see on the job. Michelle Gutierrez Vo, a registered nurse at Kaiser Permanente in Fremont and president of the California Nurses Association, says she sees the change now: "It's happening right now. And the employers are going to deny it," she said. "Employers are inserting AI technology into workflows that were usually performed, rightfully so in health care, by nurses and physicians."
A separate union, the National Union of Healthcare Workers, filed a complaint against Kaiser with state and federal authorities in July. The complaint alleges that the nonprofit health giant had implemented an algorithm to assess patients seeking mental health care, which produced recommendations for "specific levels of care without the oversight of licensed clinicians." Kaiser disputes that characterization of the tool, and the allegations have not been adjudicated.
Kaiser's response
Kaiser Permanente rejects the idea that AI makes decisions in the organization. "At Kaiser Permanente, technology, AI, and other advanced tools do not make care decisions," spokesperson Elissa Harrington wrote in emails to the Mercury News. "Our clinicians do. AI and predictive tools are designed to support clinicians and care teams, not replace their judgment."
The company points to one tool as evidence of the benefits: a predictive deterioration-alert system that Kaiser says reduced deaths by 16 percent. That figure comes from internal Kaiser research published in 2020 and is a company claim that has not been independently verified. According to the Mercury News, Harrington did not answer questions about the NUHW complaint, about the underlying technology in the mental health tool, or about the advice-nurse tool.
The industry's position
The California Hospital Association reads the law as a protection of clinical judgment, not a ban on AI support. "Clinicians should retain ultimate authority in care decisions about patients," spokesperson David Simon said, "and AI should be a supportive tool for those clinicians." His statement is reproduced only in part in the source material; according to the Mercury News, he went on to describe thoughtful, measured advancements that improve health care, without the end of his statement being quoted verbatim.
The big question: what does the law actually cover?
The law's boundaries are the murkiest part. The Mercury News reports that it is unclear whether the law prohibits any current AI use in hospitals or clinics. Bonta's spokesperson, Daniel McGreevy, provided no examples, and neither did the California Nurses Association.
Concretely, this could mean a great deal. The Wall Street Journal previously reported – reporting that the Mercury News recaps – that Kaiser had implemented an AI-based algorithm to flag patients at risk of developing sepsis, a dangerous physiological response to an infection that can lead to death, and another algorithm recommending care pathways for "advice nurses" who field calls from ill patients. Whether such tools – which offer recommendations rather than making decisions – fall on the line between support and unlicensed practice is not clarified in the public statements from either the law's author or its sponsor.
Until January 1, healthcare providers in California will have to operate with that line unresolved.
[sourceUrls]