Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Artificial Intelligence topic
No spam. Unsubscribe anytime.
Committee votes 11–1 to table bill that would restrict AI use in mental‑health services after experts warn of harms to care and research
Summary
After hours of testimony from psychologists, social workers and researchers, the House Executive Departments and Administration Committee voted 11–1 to report Senate Bill 640 inexpedient to legislate, citing concerns that the draft would hinder clinical tools and research without holding large AI platforms to account.
Get email alerts on the Artificial Intelligence topic
No spam. Unsubscribe anytime.
The House Executive Departments and Administration Committee voted to report Senate Bill 640 "inexpedient to legislate" after experts said the bill in its current form could reduce access to care and chill research on AI tools used in mental‑health treatment.
Stephanie Griffin, representing the New Hampshire Psychological Association, told the committee that the bill would "have an adverse impact on mental health in the state of New Hampshire," arguing that evidence‑based, clinician‑supervised AI tools can supplement treatment and ease access where providers are scarce. "Limiting the ability of psychotherapists to use tools like this would have an adverse impact on mental health," Griffin said during the work session.
Lynn Courier, executive director of the National Association of Social Workers — New Hampshire chapter, urged the committee to focus on unlicensed practice by platforms that effectively provide diagnoses or therapeutic communications. "If it gives a diagnosis, if it starts providing therapeutic services ... that's something that needs to be provided by a licensed professional in this state," Courier said, urging enforcement tools for the Office of Professional Licensure & Certification.
Dr. Michael Hines, a psychiatrist and assistant professor at Dartmouth, told members the bill could shut down in‑state research. He cited a similar law in Illinois that, he said, blocked a university collaboration and warned that tying carveouts to FDA approval could be premature because federal oversight of AI health tools remains unsettled.
Committee members debated whether the draft targets the companies producing large language models or instead constrains licensed clinicians who want to use regulated, HIPAA‑compliant tools under clinical supervision. Proponents of the bill described it as a stopgap to protect consumers; opponents and several witnesses said it would miss major platforms such as ChatGPT, which do not claim to offer health care and therefore would not be affected by the proposed definitions.
After the hearing and work session, Representative Slouchy moved to report the bill ITL; the motion passed on a roll call recorded in the transcript as 11–1. The committee placed SB640 on the consent calendar as "inexpedient to legislate." The committee chair said members and stakeholders should continue refining regulatory options before returning to the issue.
The committee's next procedural step is the formal committee report and placement on the House calendar; proponents and opponents said they expect further work on targeted enforcement of unlicensed practice and on a narrower approach that would regulate platforms rather than clinical use of approved tools.

