HB0452 – Artificial Intelligence Amendments

UTLawState

Date Passed

3/25/2025

Effective Date

5/7/2025

Summary

Adds targeted requirements for mental health chatbots, including clear AI disclosures, limits on selling, sharing, or using user health information and inputs for advertising, and state-filed policies describing chatbot purpose, limitations, safety protocols, and licensed-professional involvement.

Healthcare Implications

Directly affects mental-health chatbot suppliers and digital behavioral-health vendors. Covered tools need front-end AI disclosures, tighter data-use controls, advertising restrictions, and documented safety/governance policies filed with the state.

Operational Implications

  • Mental health chatbots must clearly and conspicuously disclose AI nature: before user accesses features, at start of any interaction (if user hasn't accessed in 7 days), and when user asks.
  • Suppliers may not sell or share user IIHI or user input with third parties (limited exceptions); may not use user input for advertising.
  • Suppliers must develop and file detailed policies with the state outlining the chatbot's purpose, limitations, safety protocols, and involvement of licensed mental health professionals in development.

Impact Level

High

Keywords

Transparency & Governance; Privacy & Data; Safety & Risk

Stakeholders

Patients & Public; Developers & Vendors; Providers & Health Systems; Regulators & Government