Utah Enacts AI Amendments Targeted at Mental Health Chatbots and Generative AI
Utah Updates AI laws, Focusing on Mental Health Chatbots
Updated May 31, 2025
Utah, already a leader in artificial intelligence regulation, has updated its AI laws. The state’s Artificial Intelligence Policy Act (UAIPA), enacted in 2024, mandates disclosures for consumer interactions with generative AI, especially in regulated professions like healthcare.
three new AI laws—HB 452, SB 226, and SB 332—took effect on May 7, 2025, amending and expanding the UAIPA’s scope. These laws address mental health chatbots, revise disclosure requirements for generative AI in consumer transactions and regulated services, and extend the UAIPA’s repeal date.
Mental Health Chatbot Regulations
HB 452 introduces disclosure requirements, advertising restrictions, and privacy protections for mental health chatbots. These chatbots use generative AI to simulate conversations akin to those with a licensed therapist and are presented as tools for mental health therapy or management. The law excludes AI that provides only scripted outputs like guided meditations.
A mental health chatbot must clearly disclose that it is indeed AI and not human before a user accesses its features, at the start of each interaction (if the user hasn’t accessed it in the past week), and if the user asks whether AI is being used.
Suppliers of mental health chatbots are prohibited from selling or sharing a user’s individually identifiable health information (IIHI) or user input with third parties,unless a health care provider requests it with the user’s consent,it is indeed provided to a health plan at the user’s request,or it is shared to ensure the chatbot’s functionality in compliance with HIPAA rules.
The law also restricts advertising within mental health chatbot conversations,requiring clear identification of advertisements and disclosure of sponsorships or business affiliations. User input cannot be used to determine or customize advertisements, except for ads promoting the chatbot itself.
HB 452 provides an affirmative defense for violations, contingent on creating, maintaining, and implementing a policy that meets specific requirements and filing it with the utah Division of Consumer Protection.
Violations of HB 452 may result in administrative fines up to $2,500 per violation and court action by the Utah Division of Consumer Protection.
Revised Disclosure Requirements for Generative AI
SB 226 narrows the UAIPA’s disclosure requirements for generative AI use in consumer transactions. Disclosures are now required only when an individual makes a “clear and unambiguous” request to determine if they are interacting with AI, or when generative AI is used in “high-risk” regulated services.
The law mandates disclosure when an individual clearly asks whether they are interacting with AI. It also requires regulated professionals to disclose when a “high-risk artificial intelligence interaction” occurs. This includes interactions involving the collection of sensitive personal information or the provision of personalized recommendations that could considerably impact personal decisions, such as medical or mental health advice.
A safe harbor provision protects individuals from enforcement actions if their generative AI clearly discloses that it is AI, not human, or an AI assistant at the outset of and throughout an interaction.
Violations of SB 226 may result in administrative fines up to $2,500 per violation and a court action by the Utah Division of Consumer Protection.
UAIPA Repeal Date Extended
SB 332 extends the UAIPA’s repeal date from May 1, 2025, to july 1, 2027.
What’s next
Companies offering mental health chatbots or using generative AI in Utah should assess their products and processes to ensure they comply with these new AI laws.As states increasingly regulate AI amid a deregulatory federal surroundings, healthcare companies developing and deploying AI should closely monitor state-level developments.