Lifetime Citizen Portal Access — AI Briefings, Alerts & Unlimited Follows
Senate panel reviews draft banning standalone AI mental-health services, debates penalties and oversight
Loading...
Summary
Committee reviewed a new H816 draft that would bar AI systems from providing mental-health services without a licensed professional, add a civil penalty, and require a 2027 report on regulating AI in professional practice; members asked for clarifications on definitions, HIPAA/FDA language, and enforcement.
Senate Health & Welfare reviewed a consolidated draft of H816, a bill designed to prohibit artificial-intelligence systems from delivering standalone mental-health services to Vermonters without involvement of a licensed mental-health professional. The committee did not vote and paused for a recess after a detailed staff walkthrough.
Why it matters: The draft aims to prevent psychological harm by ensuring therapeutic communications and clinical decisions remain under licensed professionals’ oversight. It also seeks to exempt supervised algorithmic tools that meet HIPAA/FDA (or other federal) requirements, while defining prohibited uses and potential civil penalties.
Committee staff explained the bill's purpose language: to "safeguard individuals taking mental health services in Vermont from psychological harm, including death by suicide, by ensuring that these services are delivered by mental health professionals and not independently by artificial intelligence systems." The draft defines therapeutic communication and therapeutic decisions, and it makes unauthorized standalone AI a potential violation of professional-conduct rules. The version the committee reviewed also incorporated a civil-penalty line (drafted as $10,000) that some staff flagged might be removed or handled differently by the Attorney General’s Office.
Members asked for clarifications on definitions (for example, whether the bill should say "licensed mental health professional" rather than a broader phrase), on the relationship between HIPAA/FDA compliance and permissible AI tools, and on the new reporting requirement: by January 15, 2027, OPR and the Board of Medical Practice would jointly report to policy committees with recommendations for regulating AI across regulated professions.
What happens next: Staff said they will refine language and follow up with OPR and AG’s office on penalty and enforcement language. Committee recessed briefly and will resume consideration with targeted questions about definitions, scope and enforcement.
