Health care coverage: prior authorization.

In a measure led by Assembly Members Harabedian and Patel, the proposal revises how utilization review and prior authorization operate for health care service plans and disability insurers, placing new timeframes and guardrails around determinations and the use of artificial intelligence in those processes. The core change sets not-to-exceed timelines for prior or concurrent requests at three business days when information is submitted electronically and five business days when submission is not electronic. For requests involving an imminent and serious threat to health, the decision window is reduced to 24 hours for electronic submissions and 48 hours for non-electronic submissions. Medi-Cal managed care plans are expressly excluded from these timeline changes, and the measure would render a willful violation a crime, create a state-mandated local program, and specify that no reimbursement is required. Implementation would also require ongoing compliance review through the Department’s periodic onsite surveys.

Key mechanisms include detailed governance of how plans and insurers review and communicate decisions. For health care service plans, the bill requires written policies and procedures for prospective, retrospective, or concurrent reviews, with criteria anchored in clinically based principles and filed with the director for review and public disclosure upon request. It adds a requirement for medical directors or clinical directors to oversee adherence to these standards, and it mandates timely communication of decisions to providers and enrollees—within 24 hours of a decision to providers, with enrollee-facing notices due within two business days for non-concurrent reviews. If information is missing or expert consultation is needed, plans must notify affected parties and provide an anticipated decision date; failure to meet timeframes can trigger administrative penalties. The legislation also introduces a comprehensive framework for the use of artificial intelligence in utilization review: AI tools must base determinations on the enrollee’s medical history, individual clinical circumstances, and other relevant records, must not rely solely on group datasets, must comply with applicable law, must not supplant human clinical judgment, and must be open to audit and public disclosure. AI systems are required to protect patient privacy, avoid discrimination, be periodically reviewed for accuracy, and permit direct clinical oversight by licensed professionals; federal guidance and regulatory developments are acknowledged, with the state empowered to issue implementation guidance and contract terms that may be exempt from certain procurement rules.

The measure situates these changes within the broader health policy landscape by aligning with existing utilization review frameworks under the Health and Safety Code and the Insurance Code, while introducing explicit protections and transparency requirements around AI-assisted decisions. It preserves Medi-Cal’s distinct timetable, emphasizes provider and patient communication and grievance mechanisms, and authorizes the state to enforce compliance through penalties administered by the director. The act’s approach reflects a policy objective to formalize performance standards for prior authorization and to establish accountable governance for AI-enabled decision-making in coverage determinations, all within the current constitutional and statutory framework that governs state reimbursements and local-government costs.