CA AB2575

Health facilities, clinics, and physician's offices using AI or clinical decision support systems for patient care must disclose detailed information to healthcare workers, including the tool's developer, training data, known biases, limitations, and validation process

Healthcare workers providing direct patient care are explicitly permitted to override AI or clinical decision support outputs when they determine an override is appropriate based on their professional judgment or to comply with civil rights law

Employers are prohibited from using technology to replace or limit a worker's professional judgment in patient care, and cannot retaliate against workers who override AI recommendations or who follow employer-approved technology outputs

Defendants who developed, modified, or deployed AI in healthcare cannot claim that a healthcare worker's failure to override an AI output is a "superseding cause" that severs the defendant's liability for patient harm

Workers subject to retaliation for overriding AI outputs may file complaints with the Labor Commissioner, and violations by health facilities constitute unfair competition under California Business and Professions Code