Which states restrict AI in therapy or mental health practice?

States such as Illinois (HB 1806), Maine (HB 2082), and Nevada (AB 406) have enacted statutes that prohibit offering therapy or psychotherapy via AI unless services are delivered by a licensed human professional. These laws generally allow clinicians to use AI for administrative drafting and supplementary workflows while barring AI from making independent therapeutic decisions or engaging in direct therapeutic communication.

Is AI therapy legal in my state?

In the United States, autonomous AI systems cannot hold a clinical license and are legally barred from practicing psychology or providing professional mental healthcare in states with enacted restrictions. Even in states without specific AI statutes, state licensing boards enforce general professional practice acts requiring human clinical oversight, informed consent, and professional judgment for all patient care.

Are mental health chatbots required to disclose that they are AI?

Yes, several states have enacted laws requiring public-facing AI chatbots and companion systems to clearly disclose their non-human status. States including California, Idaho, Nebraska, New York, Oregon, and Utah require explicit disclosure and mandate that companion chatbots implement crisis-detection protocols to refer individuals experiencing suicidal ideation to emergency services like the 988 Suicide and Crisis Lifeline.

Can an AI entity hold a health professional license?

No. States like Delaware (HB 191) and California (AB 489) expressly prohibit any non-human entity or AI agent from being licensed or certified as a healthcare professional, and bar AI systems from using professional titles or marketing terms that imply licensed clinical oversight unless a qualified human professional is actively directing the care.

What does the APA say about AI in therapy?

In its flagship Ethical Guidance for AI in the Professional Practice of Health Service Psychology (2025), the American Psychological Association emphasizes that AI should support, not replace, human professional judgment. Clinicians must maintain transparency with clients, obtain informed consent (especially when using AI scribes or documentation tools), guard against algorithmic bias, protect data privacy, and validate tools before clinical adoption.

Is there neuropsychology-specific AI guidance?

Yes. The American Academy of Clinical Neuropsychology (AACN) maintains an ongoing Disruptive Technology Initiative examining AI's role in assessment and report generation. Additionally, organizations including the International Neuropsychological Society (INS) and the American Board of Professional Psychology (ABPP) have published specialized analyses addressing generative AI workflows, test security, digital assessment, and ethical considerations specific to clinical neuropsychology.

PsychDraft

REGULATORY & CLINICAL RESOURCE HUB

AI in Psychology Practice Hub

Name: State AI Law Tracker for Psychology and Behavioral Health Practice
Creator: PsychDraft
License: https://creativecommons.org/licenses/by/4.0/

The trusted resource for clinicians tracking how artificial intelligence intersects with psychology practice, law, ethics, and professional standards.

Loading regulatory hub...