Newsom Vetoes Bill Protecting Nurses Who Override AI, Signs Clinical Judgment Law

Image source: Charles Ommanney via Wikimedia Commons, 

Part of Nurse.org’s Nursing AI Watch, our ongoing investigation into how artificial intelligence is changing the work of nurses.

If an AI tool says a patient is stable and a nurse’s assessment says otherwise, can her employer discipline her for trusting her own judgment? A California bill meant to protect her died Sept. 30, when Gov. Gavin Newsom vetoed it.

The same day, Newsom signed a separate law requiring hospitals and clinics to make sure clinicians can still use their own judgment when AI is part of patient care. The California Nurses Association backed both.

Want to see more Nurse.org articles in your Google results? Add us as a preferred source.

AB 2575, by Assemblymember Liz Ortega (D-San Leandro), covered AI tools that make predictions or recommendations about diagnosis, treatment or timing of care. It would have:

• Barred employers from punishing nurses and other direct-care workers just for following or overriding one of those tools while using their professional judgment
• Kept AI companies and health systems from escaping responsibility for a patient’s harm just by pointing out that a worker didn’t override the tool

The bill failed in the state Senate on Aug. 28, then passed on a second try on the last night of the session.

Newsom praised the goal, but his veto message said the bill wasn’t “practical and enforceable” as written. He named two problems:

• The protection was too narrow. It covered only punishment based solely on an override, a higher bar than other retaliation laws. Newsom said that “ties the Labor Commissioner’s hands in establishing a violation.”
• A labor office could be pulled into judging patient care. Newsom argued the bill would put the Labor Commissioner in the position of deciding whether an override met the standard of care, “something the Labor Commissioner does not have the medical expertise or skill to do.” The bill itself said the commissioner would not have to make a final call on that question.

>>Listen to The Latest Nurse News Podcast

AB 1979, by Assemblymember Mia Bonta (D-Oakland), takes effect Jan. 1, 2027. The new law requires hospitals, clinics and physician offices to:

• Take reasonable steps so licensed clinicians can still use their own judgment when AI informs care
• Stop using AI to direct unlicensed staff in work that requires a license, or to do that work itself

It also puts consumer health chatbots under California’s medical privacy law.

Earlier drafts flatly banned using AI to replace a clinician’s judgment. Lawmakers softened that to “reasonable steps,” and the California Hospital Association went neutral on the amended bill. What the law does not do is protect a nurse from discipline for overriding AI. That was AB 2575’s job.

How Nurses and Hospitals Reacted to the Veto

“Today, a nurse who sees that the algorithm is wrong and acts to save her patient can still be disciplined or fired by her employer,” said Sandy Reding, RN, a president of the California Nurses Association. In a 2024 National Nurses United survey of more than 2,300 RNs, 69% of those working with acuity algorithms said the output didn’t match their own assessments of their patients.

Nurse.org’s own 2026 State of Nursing survey found the same doubts nationally: just 22% of nurses trust AI tools to support safe patient care, and 40% say nurses have no meaningful input into how those tools are chosen.

The California Hospital Association, which urged a veto, argued the bill would have kept hospitals from correcting a worker even when an override harmed a patient. The final bill was narrower: it covered only punishment based solely on following or overriding an AI tool, and it kept nurses’ existing duty to meet the standard of care.

Backers plan to try again. The California Federation of Labor Unions, which co-sponsored the bill, said it is “committed to bringing this bill back next year,” and Ortega said she “will keep fighting.” Any new version will go to a new governor, since Newsom’s term ends in January 2027.

Related Nursing AI Watch Analysis

🤔 Have you ever overridden an AI alert or acuity score at work, and what happened next? Tell us in the comments below.