Colorado Legislates AI Guardrails in Healthcare, Mental Health, and Insurance

Colorado is at the forefront of state-level AI regulation, introducing two significant bills aimed at governing the use of artificial intelligence within its healthcare system. House Bill 1195 specifically targets mental health, prohibiting licensed therapists from engaging in direct chatbot communication with patients. It also mandates human approval and review for any treatment plans or therapeutic recommendations generated by AI, ensuring that core psychological care remains human-led. Furthermore, the bill requires explicit written informed consent from clients if AI is used to record or transcribe therapy sessions, addressing privacy and trust concerns. The legislation also restricts companies from marketing AI tools as psychotherapy, emphasizing that such services must be delivered by regulated professionals.

The second piece of legislation, House Bill 1139, focuses on health insurance. It explicitly bans insurance companies from using AI systems as the sole basis for denying coverage. This bill requires that individual medical histories and circumstances be considered in coverage decisions, and that all denials undergo review by a qualified human expert. Both bills share common ground in promoting transparency, requiring clinicians to inform patients precisely when and how AI is being utilized in their care, and strictly prohibiting AI chatbots from misrepresenting themselves as human licensed clinicians.

While both proposals garnered significant support, with HB 1195 passing committee unanimously, industry groups like the Colorado Technology Association have expressed concerns. They advocate for clear guardrails but suggest that some provisions might be overly broad or difficult to implement, potentially restricting responsible AI applications. The legislative sponsors, however, emphasize that the intent is not to ban AI but to establish clear 'rules of the road' for its ethical and transparent deployment, particularly where it intersects with sensitive patient data and critical health decisions. These bills represent a crucial step in defining the boundaries of AI in healthcare, prioritizing human oversight and patient well-being.