Summary
Presents a risk-based framework and twelve policy considerations for AI in mental health and substance use services, including independent evaluation, vendor transparency, premarket and postmarket oversight, crisis and adverse-event reporting, disclosure, age verification and accountability.
Healthcare Implications
Provides nonbinding considerations for policymakers and behavioral health leaders governing administrative, clinician-facing and patient-facing AI. Proposals for reporting, safety testing, disclosure and recourse are recommendations for policy development, not newly imposed legal duties or authorization for autonomous therapy.