The legislative landscape is shifting rapidly. California's SB 1120 (effective 2025) requires a licensed physician to make final medical necessity determinations. Washington's SB 5395 mandates that AI review criteria include individual patient data and state audits for fairness. Nebraska's LB 77 bars AI output as the sole basis for denials and requires disclosure of AI use. Texas's SB 815 limits AI to administrative or fraud detection, prohibiting it in adverse determinations. Colorado's HB 26-1139 requires licensed clinician review of AI-generated denials and regulatory disclosure. Over 47 states introduced more than 250 AI healthcare bills in 2025 alone.
The regulatory push reflects federal pressure and industry concern. The Centers for Medicare and Medicaid Services clarified in 2024 that algorithms cannot deny admission or terminate care without patient-specific reassessment, and Medicare Advantage regulations require decisions based on individual circumstances reviewed by qualified professionals. States accelerated their own protections after early drafts of proposed federal legislation threatened preemption. For insurers and healthcare counsel, Utah's January 2027 compliance date signals imminent operational changes: disclosure requirements, human review protocols, and potential liability exposure for algorithmic denials now face statutory enforcement across multiple jurisdictions.