Two bills under consideration would regulate chatbots and restrict AI in psychotherapy

By Mona Abou, for the Community News Service

As artificial intelligence proliferates, Stephanie Winters is concerned for Vermonters’ protections and privacy in health care. 

“The rapid development of artificial intelligence and neurotechnology raises legitimate questions about privacy, autonomy, transparency and protection from misuse,” Winters, the deputy executive director of the Vermont Medical Society and Vermont Psychiatric Association, told the House Committee on Health Care on Feb 27.

Lawmakers are weighing two bills – H.814 and H.816 – that would create a new set of protections from neurotechnology in Vermont and regulate usage of artificial intelligence in mental health services, respectively. 

The first bill, H.814, would regulate chatbots that act as stand-ins for mental health professionals; require disclosures when generative AI is being used in health care settings; and restrict how health plans use AI in evaluating medical care. 

The bill, which has 14 sponsors, also includes provisions to protect health data when using “brain-computer interfaces,” which use neural signals to control devices such as robotic limbs. Patients would have to consent to share their data and be able to revoke it at any time, after which their records would be destroyed within 10 days.

H.814 would also restrict health care insurers from using AI to deny, delay or modify a customer’s health care based on medical information, instead deferring to a licensed human health care provider. Additionally, Vermont’s Artificial Intelligence Advisory Council would have to report on ethical and responsible AI use in health care, human services and education to lawmakers by next January. 

H.814 goes hand in hand with H.816, which would regulate how mental health professionals use AI. More specifically, it would prohibit them from advertising mental health services that use AI to provide therapeutic judgement, diagnosis or treatment. 

Lynn Currier, executive director of the National Association of Social Workers, testified in support of H.814, saying that it would help Vermont set standards around an emerging technology. She raised concerns about the capabilities of AI therapy chatbots in Vermont and wants them banned.

“At best, it’s unlicensed practice. At worst, anything that’s attached to a large language model is potentially very dangerous,” Currier told the House Committee on Health Care on Feb. 25.

Dr. Rick Barnett, a licensed psychologist who chairs the Legislative Committee for the Vermont Psychological Association, said the organization supports H.814. But he cautioned lawmakers about restricting chatbots because the U.S. Food and Drug Association encourages their use. 

“We don’t want to get dinged, violating these new laws for using an FDA approved product,” Barnett said during the Feb. 25 meeting.

Winters, of the Vermont Medical Society, said she agrees with the H.816’s goal to ensure AI doesn’t replace licensed professionals, but she advised lawmakers to tighten the wording so it doesn’t ban helpful tools. 

“The practical consequence could be increased administrative burden and reduced adoption of tools designed to reduce clinician burnout,” Winters said. “H.816, as written, does not currently include an explicit supervised use exception clarifying that AI assisted outputs are permissible if reviewed and approved by a licensed professional.”

Lawmakers are continuing to hear testimony on the bills and have not yet voted on either of them.

Via Community News Service, a University of Vermont journalism internship