AI in Health Care: Tennessee and North Carolina Show New AI Rules for Health Care

In Tennessee and North Carolina, AI regulation is expanding beyond privacy and recordkeeping issues. Lawmakers are now looking at patient-facing claims, behavioral health tools, chatbots and coverage-review processes.

Tennessee has passed a law that bars people who develop or use AI from claiming that an AI system is, or can act as, a qualified mental health professional. North Carolina lawmakers considered bills that would have required notice to patients when AI is used to make medical-necessity decisions and meaningful human review of AI-derived coverage decisions. Those bills did not move forward.

For physicians and health care leaders, the message is clear: review how AI tools are described to patients, make sure chatbots and digital tools do not claim too much, and build human review into your key workflows before lawmakers require it.

Tennessee Limits AI Mental Health Claims

Senate Bill 1580 / House Bill 1470 (effective July 1, 2026) bars any person who develops or uses AI from advertising or telling the public that the AI system is, or is able to act as, a qualified mental health professional. The law affects behavioral health providers, hospitals, clinics, telehealth platforms, app developers, chatbot vendors and organizations using AI-enabled mental health support tools.

For health care groups, the main issue is not whether they can use AI. Rather, the main issue is how they describe it to patients. Patient webpages, chatbot openings, intake forms, scripts, app descriptions, ads and vendor content should not suggest that an AI tool is a therapist, counselor, psychologist, psychiatrist, or a substitute for a licensed professional. Organizations should also train staff to describe AI tools as aids that support care, not as providers of mental health care.

North Carolina Eyes AI Rules for Health Care

North Carolina lawmakers considered two bills during the 2025-2026 legislative session. While neither bill passed, both show where new AI rules may emerge. Senate Bill 624 would have required providers to tell patients when AI is used to decide whether a service or treatment is needed. Also, the bill would have placed rules on AI health chatbots. Senate Bill 316 would have required licensed health professionals to review AI-based coverage decisions tied to prior authorization requests before those decisions could be used.

These proposals point to three areas lawmakers continue to watch:

• patient notice
• chatbot oversight 
• human review of AI-supported coverage and care decisions

What Should Health Care Groups Do Now? 

Organizations operating in Tennessee and North Carolina should focus on patient communications, behavioral health AI tools, chatbot design and human review.

• Behavioral health operations. Providers using AI for mental health, wellness, triage, intake, or patient engagement should review disclaimers, chatbot scripts, website language, staff talking points and other patient-facing materials.
• Human review. Health care groups should consider adding review steps now so that AI-generated recommendations, chatbot outputs, prior authorization decisions and coverage decisions do not become final without clinical review.
• Marketing and patient communications. In Tennessee, AI tools should not be marketed or described as mental health professionals or as replacements for licensed clinicians.
• North Carolina readiness. Even though the bills did not pass, they point to future areas of regulation including patient notice, oversight of AI health chatbots and human review of AI-based coverage decisions.
• Vendor controls. Vendors often supply the language that patients see. Health care providers should review vendor scripts, prompts, user interfaces, marketing materials and disclaimers before using them.

What are the Key Steps to Take Now?

Consider the following:

• Audit patient-facing language. Review websites, ads, chatbot introductions, intake forms, app descriptions, marketing materials and call-center scripts for statements that suggest AI can provide therapy or can act as a licensed mental health professional.  Chatbots and other digital tools should clearly communicate that they support care and they do not replace licensed professional judgment, diagnosis, treatment or emergency care.
• Build human review into workflows. Set policies that require licensed professionals to review AI-generated clinical recommendations, patient communications and coverage-related information before anyone relies on them or presents them as clinical conclusions.
• Control vendor content. Require vendors to submit patient-facing materials for review and to update scripts, prompts, labels and marketing materials when state rules change. Providers should review and approve all vendor content before using it.
• Create an AI Inventory. Review current AI tools, identify where AI affects patient communications or coverage workflows and confirm that licensed professionals remain meaningfully involved in clinical and coverage-related decisions.
• Track legal changes. Assign responsibility for tracking Tennessee enforcement activity and North Carolina AI bills so that policies, training and vendor requirements can be updated as needed.
• Train Staff. Train clinicians, front-desk teams and marketing staff to describe AI mental health tools correctly. These are tools that support licensed professionals rather than replace  human clinical judgment.

Please contact Beau Haynes, Courtney Hurtig, Matt Harrell, Lucy Porter or any member of the Health Care or Artificial Intelligence teams to discuss how these developments may affect your operations and how to build a compliance roadmap.