transparencycoalition.ai web signal

Rhode Island Signs Three AI Laws Covering Mental Health Chatbots

TL;DR

  • Governor Dan McKee signed three AI laws on June 22, 2026, targeting mental health chatbots, AI companions, and healthcare documentation.
  • Chatbot operators face up to $15,000 per day in fines for failing to route users expressing self-harm to crisis services.
  • A ban on AI simulating emotional attachment prohibits unlicensed AI companions marketed for mental health or emotional support.

Rhode Island's governor signed three AI laws in a single session, and the package is more specific than the usual 'AI guardrails' announcement. According to the AI Transparency Coalition, Governor Dan McKee signed the trio on June 22, 2026, addressing three distinct pressure points: crisis response by chatbots, the AI therapy and companion market, and AI documentation in clinical settings.

The first law requires chatbot operators to build in protocols that detect expressions of suicidal ideation or self-harm and route users to crisis services: suicide hotlines, crisis text lines, and similar resources. Starting July 1, 2027, operators must file annual reports with the state attorney general documenting how often those protocols activated, with the data published in aggregate. Noncompliance carries penalties of up to $15,000 per day, with fines directed back to suicide prevention programs.

The second law is broader and more consequential for the AI mental health market. It bans unlicensed individuals and businesses from offering therapy or psychotherapy via AI, and goes further to prohibit AI from simulating emotional attachment, bonding, or dependency. AI companions marketed for mental health or emotional support are specifically called out, as is AI making independent therapeutic decisions or determining treatment plans.

The third law covers clinical documentation: healthcare providers using AI to document in-person or telehealth visits must notify patients and review the AI-generated records for accuracy.

The honest caveat is that several provisions will hit edge cases quickly. The prohibition on simulating emotional attachment is broad enough to reach mainstream consumer apps with conversational features, and enforcement against out-of-state operators is a gap the reporting does not address. For clinical AI documentation vendors with existing audit trails and notification features, the third law creates a compliance checkpoint that favors established players. For the broader AI therapy and companion market, Rhode Island's specificity with named behaviors and named penalty structures gives a clearer legislative signal than most state AI bills have offered so far.