US States Rush to Regulate AI Chatbots: Rhode Island Signs Therapy Bot Ban
Summary: As federal AI legislation stalls, US states are filling the regulatory vacuum — particularly in mental health and education. Rhode Island signed three AI laws on June 22, while California and Arizona took divergent approaches.
Key Points
- Rhode Island: Gov. McKee signed three AI bills — (1) ban on AI systems offering therapy without a licensed professional, penalties up to $15K/day with fines directed to suicide prevention programs; (2) mandatory crisis-service routing when chatbots detect suicidal ideation or self-harm; (3) patient notification requirement when AI transcribes clinical visits
- California: Legislature passed a ban on AI-only public school teachers and sent it to Gov. Newsom for signature
- Arizona: Gov. Hobbs vetoed all three AI bills the legislature passed — the most permissive stance among governors this month
- 78 chatbot-related bills are currently moving through legislatures in 27 US states
Why It Matters
Without comprehensive federal AI law, a fragmented patchwork of state rules is taking shape. Rhode Island's licensed-professional-plus-fines model is likely to be replicated elsewhere, starting with states where mental health chatbot harms have made headlines. If AI restrictions spread from mental health and education to other sectors, companies will face 50 different compliance regimes — a significant operational burden that could reshape how AI services are offered in the US.
Further Reading
- Rhode Island AI Laws Detail — Transparency Coalition
- June 26 Legislative Update — Transparency Coalition