AB-2575 Health care services: artificial intelligence

What happened

AB 2575’s retained proposal would prohibit employer retaliation or discrimination based solely on a direct-care worker’s override of, or reliance on, clinical decision-support output when exercising professional judgment within scope. It preserves duties concerning the standard of care, independent judgment, licensure and applicable policies, while excluding a policy that demands deference to the system from the protected policy duties. The Labor Commissioner could review retaliation complaints, without deciding professional negligence or licensing discipline.

A separate proposed civil-liability provision would bar a system developer, modifier, selector or deployer from asserting that a worker’s failure to override the output was a superseding cause severing liability. It preserves other affirmative defenses, relevant causation and foreseeability evidence, comparative fault and reasonable-precaution evidence. It excludes specified professional-negligence actions against licensed or certified direct-care professionals. These distinctions do not automatically establish a defendant’s liability or excuse clinical responsibilities. The measure remained pending at the September 19 cutoff.

Why it matters

Clinical AI policies and vendor agreements involve both employment and liability questions, with different standards and exceptions.

Sources

• AB-2575 Health care services: artificial intelligence, bill historyOfficial · California Legislature · 2026-08-31 · Source review 2026-09-20
• AB-2575 Health care services: artificial intelligence, current bill text (09/04/26 - Enrolled)Official · California Legislature · 2026-09-04 · Source review 2026-09-10