Summary
States that generative AI and large language models must not be represented as physicians or licensed mental health professionals or independently make treatment or prescribing decisions. Calls for licensed-professional governance, clear non-clinical labeling, human-care and crisis pathways, appropriate disclosures, and protection of health information.
Healthcare Implications
Mental health providers and developers should preserve clinician responsibility, avoid marketing chatbots as clinicians or therapy, disclose system limitations, obtain appropriate consent for health-data uses, and provide access to human and emergency support.