Summary
Creates a broad Vermont health-AI law covering neurological rights, neural-data privacy, generative-AI patient communication disclosures, mental health chatbot privacy/disclosure/governance requirements, and health-plan use of AI, algorithms, or software tools in utilization review and utilization management. Also extends and expands the Artificial Intelligence Advisory Council’s health, human services, education, and utilization-review reporting responsibilities.
Healthcare Implications
Directly affects providers, mental health chatbot suppliers, health plans, patients, and state regulators. Health care providers must label unreviewed generative-AI clinical communications and provide human-contact instructions; mental health chatbot suppliers must disclose AI use, limit use and sharing of user health information, maintain safety/governance documentation, involve licensed mental health providers in development/review, and file policies with the Attorney General to claim an affirmative defense; health plans cannot use AI to deny, delay, or modify medically necessary services and must use individual clinical context, fairness safeguards, auditability, and licensed human review.
Operational Implications
- Health care providers using generative AI to generate written or verbal patient communications relating to patient clinical information must include a disclaimer that the communication was AI-generated and instructions for contacting a human health care provider or appropriate employee; communications reviewed by a licensed human provider are exempt.
- Mental health chatbot suppliers must clearly and conspicuously disclose that the chatbot is AI and not human before a Vermont user accesses the chatbot, at the beginning of an interaction after seven days without access, and whenever the user asks whether AI is being used.
- Mental health chatbot suppliers may not sell or share a Vermont user's individually identifiable health information or user input except in limited provider, health plan, user-requested, or contracted-service circumstances; health plan AI patient data also may not be used beyond its intended and stated purpose.
- Mental health chatbot suppliers seeking the statutory affirmative defense must include procedures for assessing and responding to user harm, responding in real time to acute physical-harm risk, letting users report harmful interactions, mitigating foreseeable adverse outcomes, and prioritizing user mental health and safety over engagement metrics or profit.
- Mental health chatbot suppliers seeking the affirmative defense must create and implement written policies and maintain documentation on foundation models, training tools, federal health privacy compliance, data collection and sharing, accuracy, reliability, fairness, safety, clinical best-practice monitoring, pre-release and regular testing, audits/reviews, and anti-discrimination measures; health plans using AI must periodically review and revise performance, use, and outcomes to maximize accuracy and reliability.
- Mental health chatbot suppliers seeking the affirmative defense must file their written policy, supplier and chatbot name/address, and filing fee with the Office of the Attorney General; health plans must include required disclosures on AI use and oversight in written policies and procedures to the extent required by the Department of Financial Regulation.
- Health plans using AI, algorithms, or other software tools for utilization review or utilization management based in whole or part on medical necessity may not use those tools to deny, delay, or modify health care services; medical-necessity determinations must be made only by a licensed human health care provider competent to evaluate the specific clinical issues.
- Health plan AI tools must base determinations on the covered individual's medical or clinical history, the specific clinical circumstances presented by the requesting provider, and other relevant clinical information in the covered individual's record, and may not rely only on a group dataset.
- Health plan AI tools must not discriminate directly or indirectly against covered individuals in violation of state or federal law and must be fairly and equitably applied, including in accordance with applicable HHS regulations and guidance.
- Health plan AI tools must be open to inspection, audit, and compliance review by the Department of Financial Regulation and other state agencies/departments under applicable law.
- The Artificial Intelligence Advisory Council and Division of Artificial Intelligence must report to the General Assembly by January 15, 2027 on additional statutory changes, including neurological-rights protections, guidance on generative AI by regulated professions, and regulation of artificial or augmented intelligence in health insurance utilization review.