An act relating to artificial intelligence
H.821 is an artificial intelligence regulation and oversight bill that expands Vermont’s Artificial Intelligence Advisory Council, extends its sunset date, and directs it to study the effects of AI on education, finance, public health, human services, and public participation. The bill also requires the Council, working with the Director of the Division of Artificial Intelligence, to review professional guidance, research public-sector uses of AI, and report to the General Assembly by January 15, 2027 with recommendations for ethical and responsible AI use and possible pilot projects.
The bill creates new consumer disclosure requirements for generative artificial intelligence. A person using generative AI to interact with a consumer must disclose, when asked or prompted, that the consumer is interacting with AI rather than a human. For regulated occupations—such as licensed health, legal, financial, and mental health services—the bill requires prominent disclosure when generative AI is used in a high-risk interaction involving sensitive data or advice that could affect significant personal decisions. It also makes violations an unfair and deceptive act in commerce, gives the Attorney General enforcement authority, and bars a defendant from using AI as an affirmative defense to a consumer-protection violation.
The bill would amend 3 V.S.A. § 5023 to change the membership of the Artificial Intelligence Advisory Council, add representatives with expertise in education, health care, civil liberties, ethics, law, racial equity, and finance, and extend the council’s repeal date from 2027 to 2030. It would also add a new chapter to Title 9 establishing statewide disclosure rules for generative AI in consumer interactions and regulated professional services, and it would create a new consumer-protection rule that AI cannot be used as a defense to a violation of Vermont’s unfair and deceptive acts statute. The Attorney General would gain rulemaking, investigative, and enforcement authority over these AI disclosure provisions.
The bill appears generally supportive of stronger AI oversight, consumer transparency, and ethical use of AI in public and professional settings. Its structure suggests a precautionary approach: it does not ban generative AI, but instead seeks to study its impacts, require disclosure, and preserve existing consumer-protection standards. No committee transcript or vote record was provided, so there is no recorded legislative debate or vote sentiment to assess beyond the bill’s text and stated purpose.
The main points of potential contention are likely to be the scope of disclosure obligations, especially for businesses and licensed professionals using generative AI, and the breadth of what counts as a “high-risk artificial intelligence interaction.” Regulated occupations may object to the compliance burden or to the requirement that AI use be prominently disclosed in sensitive service settings, while consumer advocates and civil liberties groups are likely to support the transparency requirements. Another possible point of debate is the bill’s expansion of the council and extension of its sunset date, which may raise questions about whether the advisory body should remain temporary or become a longer-term fixture.