Summary
Strengthens Washington’s prior-authorization framework by adding explicit guardrails on AI and automated decision systems used in medical-necessity and coverage determinations, including human clinical review, disclosure, auditability, periodic review, individual-context requirements, non-discrimination standards, and limits on secondary patient-data use.
Healthcare Implications
Directly constrains payer and utilization-management use of AI in coverage decisions. Health plans and UR vendors must maintain qualified human review, document and disclose AI use, support audits, periodically review AI performance, consider individual clinical context, and limit patient data to its intended and stated purpose.
Operational Implications
- AI may not be the sole basis for adverse medical-necessity determinations; clinician must review.
- AI must account for enrollee's individual clinical circumstances, not rely solely on group data.
- AI tools must be applied equitably and must not discriminate.
- Commissioner has audit authority over insurer AI use.
- AI performance, use, and outcomes must be periodically reviewed.
- Insurers must disclose AI use in UR to regulators, providers, and enrollees.
- Patient data used in AI may not be used beyond the intended and stated purpose.