Chatbots provide many opportunities for innovative health services, such as diet and exercise recommendations, insights from wearable devices, mental health tips, guided meditation, and administrative support (e.g., scheduling appointments and assisting with insurance questions). Companies in this space will need to consider legal and regulatory risks, especially when services are accessible to children and teens, or could appear to provide professional services in the U.S.

Increasing federal and state scrutiny

In 2025, the Federal Trade Commission, FDA, U.S. House and Senate members, and state attorneys general led hearings, sent letters, and launched investigations regarding AI chatbot safety. In January 2026, the Kentucky state Attorney General brought the first enforcement action against a major companion chatbot provider, alleging, among other things, that some of the provider's chatbots are providing minors with mental health advice without a professional degree.

Rapidly evolving legal landscape

Lawmakers are quickly pursuing AI chatbot safety legislation. Some state laws impose requirements on "companion" chatbots, such as protocols for suicidal ideation and disclosures that a system is "non-human." Other state laws – including those in California, Illinois, Nevada, and Utah – restrict or prohibit the non-licensed provision of mental health services. Looking ahead, state and federal lawmakers are considering a range of new bills, including a federal bill that would address AI chatbots that appear to provide any type of professional service.

Legal and regulatory risks

Companies will need to consider numerous risks as they integrate AI chatbots into health services, to determine whether the services could trigger statutory obligations or lead to investigations or lawsuits. For example, it will be important to assess whether an AI chatbot seems to provide a professional service (e.g., examining factors such as marketing materials, user experiences, or outputs that might imply clinical authority or therapeutic support), and whether notice that a chatbot is being used or other measures may be required or help mitigate risk. Companies may also wish to consider whether "companion" chatbot laws could apply due to features that simulate a humanlike relationship, such as conversational or emotionally responsive AI systems. Additionally, any use case that is accessible to children and teens is likely to draw scrutiny from lawmakers and regulators in 2026.