transparencycoalition.ai web signal

Vermont Gov. Phil Scott Signs H.816 Therapy Chatbot Ban

regulation safety ai-ethics ai-policy regulation mental-health consumer-protection

TL;DR

  • Vermont Governor Phil Scott signed H.816 on June 17, prohibiting the use of therapy chatbots and regulating AI in mental health services.
  • Colorado Governor Polis signed a similar psychotherapy AI restriction, HB 1195, two weeks earlier on June 3.
  • Missouri's SB 1019 also includes a prohibition on offering AI therapy chatbots, extending the multi-state pattern.

Vermont Governor Phil Scott signed H.816 into law on June 17, adding the state to a growing cluster of jurisdictions drawing explicit lines around AI in mental health contexts. The bill, sponsored by Representatives Berbeco, Arsenault, and Priestley, prohibits the use of therapy chatbots and regulates AI in the provision of mental health services, according to the Transparency Coalition's weekly AI legislative update.

Vermont has been described in some reporting as the first U.S. state to act here, but the Transparency Coalition's update makes no such claim — and the timeline doesn't support it. Colorado Governor Polis signed HB 1195, which covers AI restrictions related to psychotherapy services, on June 3, two weeks before Vermont. Missouri's SB 1019 also includes a prohibition on offering AI therapy chatbots, though the article doesn't detail its current status. Vermont is part of a wave, not the front of a parade.

That context matters for how product teams should read this. Mental health and wellness AI products now face a patchwork of state-level restrictions signed within weeks of each other, with potentially different scopes. Whether H.816 applies only to platforms that explicitly market themselves as therapy tools, or reaches coaching and wellness apps that stop short of that language, is a question the article does not answer, and the distinction will matter a great deal in practice.

The honest caveat is that the Transparency Coalition's update is a tracking digest, not a legal analysis. It confirms the signature and the general category of restriction but provides no detail on enforcement mechanisms, penalties, or exactly what threshold a product must cross to fall under the ban. What the reporting doesn't give you is any comparison of Vermont's specific statutory language to Colorado's, or clarity on whether a product compliant in one state would pass muster in the other.

For builders, the forward-looking read is straightforward: two therapy AI restrictions signed within two weeks, Missouri pending, is pattern evidence rather than coincidence. Companies deploying in mental health or wellness-adjacent categories now need multi-state compliance reviews, not just attention to whichever headline crossed their feed first.