Status: Enacted — Signed by Governor Phil Scott on June 17, 2026 as Act 156. Effective on passage.
Bill Number: Vermont H.816 (2025-2026 Session) — An act relating to the use of artificial intelligence in mental health services (Act 156)
Jurisdiction: Vermont
Category: AI Governance, Healthcare AI, Mental Health AI, Professional Licensing, Consumer Protection
Industries: Behavioral Health, Telehealth, Digital Mental Health Apps, Consumer AI
Enacted Date: June 17, 2026
Effective Date: On passage (June 17, 2026)
Summary: Vermont H.816, enacted as Act 156, imposes some of the strictest state-level restrictions in the United States on the use of artificial intelligence in mental health care. The law prohibits corporations and other entities from providing mental health services independently through AI systems, bars AI from making mental health diagnoses or treatment plans, and prohibits AI from offering therapeutic guidance. It preserves practical, non-clinical uses such as scheduling, billing, and transcription that support — rather than replace — licensed clinical judgment. Signed by Governor Phil Scott on June 17, 2026, the law treats AI-enabled therapeutic decision-making as unprofessional conduct and ties enforcement to Vermont's Consumer Protection Act. It follows Tennessee's AI in Mental Health Care Restriction (SB 1580) and parallels a national surge of "therapy bot" measures, including Illinois SB 3114, Colorado HB 26-1195, Nevada and Utah restrictions, and bills advancing in Rhode Island and Pennsylvania during the 2026 sessions.
Key Requirements:
- Prohibition on AI-Only Therapy: Corporations and entities may not provide mental health services independently through an AI system; therapeutic services must be delivered by a qualified, licensed professional.
- No AI Diagnoses or Treatment Plans: AI systems are prohibited from making mental health diagnoses or generating treatment plans.
- No AI Therapeutic Guidance: AI may not offer therapeutic guidance directly to individuals in place of a licensed clinician.
- Permitted Administrative Uses: The law preserves AI for supporting functions such as scheduling, billing, and transcription that do not substitute for clinical judgment.
- Enforcement: AI-enabled therapeutic decision-making is treated as unprofessional conduct, and enforcement runs through Vermont's Consumer Protection Act, giving the Attorney General authority to act. Vermont's AI Advisory Council must submit a mental health AI regulation report by January 15, 2027.
Legislative Context: H.816 was sponsored by Representatives Daisy Berbeco, Angela Arsenault, and Monique Priestley, and was one of two AI bills Governor Scott signed in mid-June 2026 (alongside H.211 on data brokers). The therapy-bot restriction reflects growing concern about unlicensed AI products marketing themselves as mental health support, particularly to vulnerable users. It situates Vermont among the most active states limiting clinical AI, joining a cluster of mental-health AI statutes enacted across the country in 2026.
Affected Entities: Companies offering AI-driven mental health or "therapy bot" products to Vermont residents, telehealth and behavioral-health providers integrating AI into clinical workflows, and digital mental health app operators. The law governs clinical and therapeutic AI; it does not reach physical security systems, video analytics, or computer-vision safety platforms.
Compliance Outlook: Behavioral-health and digital mental health operators serving Vermont should audit any AI feature that could be construed as diagnosing, treating, or providing therapeutic guidance, and route those functions through licensed clinicians. Marketing claims positioning an AI product as a substitute for professional care carry particular risk under the Consumer Protection Act framing. Because Vermont, Tennessee, Illinois, Colorado, and others converge on the principle that AI may assist but not replace licensed mental health professionals, multi-state operators should standardize on a "human clinician required" design. The law has no operational impact on physical security or visual-analytics platforms.
IntelliSee Relevance: Adjacent — Healthcare Sector Context. Act 156 regulates clinical and therapeutic AI in mental health care; it does not regulate physical security AI, video monitoring, or weapons detection. IntelliSee's AI analyzes live camera feeds for physical safety events and has no diagnostic, therapeutic, or conversational function, so the law imposes no direct compliance obligation on IntelliSee deployments. It is tracked because IntelliSee serves healthcare and behavioral-health facility security teams — hospital and clinic security directors and risk managers — who operate in Vermont's tightening AI-governance environment. Healthcare customers deploying IntelliSee for AI gun detection, slip-and-fall detection, and aggressive-behavior monitoring should note these physical-safety tools sit entirely outside the clinical-AI regulatory perimeter. See the State-by-State AI Security Legislation Tracker for full context.
Compliance Timeline
- January–March 2026 — H.816 introduced and passed by the Vermont House.
- May 2026 — Senate amendments and conference committee report adopted by both chambers.
- June 17, 2026 — Governor Phil Scott signs H.816 into law as Act 156; effective on passage.
- January 15, 2027 — Vermont AI Advisory Council mental health AI regulation report due.