AI Regulation Tracker / State law
Illinois WOPR Act bars AI from therapy and limits it to admin support, with fines up to $10,000
Illinois has made it unlawful to offer therapy unless a licensed professional delivers it, and it forbids AI from therapeutic communication, treatment planning, and emotion detection. The state can fine violators up to $10,000 per incident.
Illinois has drawn a clear legal line around artificial intelligence in mental health. The Wellness and Oversight for Psychological Resources Act, filed as HB 1806 and enacted as Public Act 104-0054, was signed by Governor JB Pritzker and took effect on August 1, 2025. Its rule is direct. A person, corporation, or entity may not provide, advertise, or otherwise offer therapy or psychotherapy to the public in Illinois unless a licensed professional conducts those services. Software cannot stand in for a clinician, and a company cannot market an AI product as if it could.
The Act was sponsored by Representative Bob Morgan, with Representative Michael Crawford as chief co-sponsor. It enforces its therapy-provider requirement through a companion set of restrictions on how licensed professionals themselves may use AI. Those restrictions are the part most behavioral-health operators need to read closely, because they define the boundary between a permitted tool and a prohibited substitute.
What AI is barred from doing
The statute is specific about the clinical functions AI may not perform. A licensed professional may not use an artificial intelligence system in therapy or psychotherapy services to make independent therapeutic decisions, to directly interact with clients in any form of therapeutic communication, or to generate therapeutic recommendations or treatment plans without the review and approval of a licensed professional. The Act separately forbids the use of AI to detect emotions or mental states.
Read together, these prohibitions target the core of clinical practice. The judgment call, the therapeutic conversation, the plan of care, and the reading of a client's emotional condition all have to run through a human clinician. An AI that drafts a treatment plan is not automatically unlawful, but a plan that reaches a client without a licensed professional reviewing and approving it is. The emotion-detection ban is the most categorical of the four. There is no review-and-approval exception attached to it in the statutory language, which means affect-recognition features aimed at inferring a client's mental state sit outside the permitted zone regardless of human oversight.
What AI is still allowed to do
The Act does not outlaw AI in a therapy practice. It confines it. Licensed professionals may use AI for administrative and supplementary support, subject to consent and disclosure conditions. In practice that covers the back-office layer of a modern practice: scheduling, appointment reminders, billing, transcription and note formatting, and other clerical work that does not involve clinical decision-making or direct therapeutic interaction. The distinction the legislature drew is between AI that assists a clinician's operations and AI that performs the clinician's job. The first is allowed. The second is not.
That line matters for product design. A documentation assistant that a therapist reviews and signs off on looks different, legally, from a chatbot that talks a client through a panic episode. The first is supplementary support. The second is therapeutic communication, and it is barred.
Who enforces it and what it costs
Enforcement sits with the Illinois Department of Financial and Professional Regulation, the same agency that licenses the state's mental-health professionals. The Act authorizes civil penalties of up to $10,000 per violation, assessed with reference to the degree of harm and the circumstances of the violation. The per-violation structure is significant for platforms operating at scale. A single noncompliant feature deployed across many client interactions is not necessarily one violation, and the penalty ceiling applies to each violation rather than to a course of conduct as a whole.
Why this is not the Illinois employment-AI law
Illinois passed more than one AI measure in this period, and the two are easy to confuse. HB 1806 governs therapy and behavioral health. It is distinct from HB 3773, the amendment to the Illinois Human Rights Act that addresses AI in employment decisions and takes effect on its own timeline. A compliance team that has already worked through the employment-AI changes has not, by doing so, addressed the WOPR Act. The obligations, the covered activity, and the enforcing framework are different, and behavioral-health providers should treat the WOPR Act as a separate workstream.
The reach beyond Illinois
Illinois is not the only jurisdiction acting on AI and mental health, and platforms rarely serve one state. Any teletherapy or AI-wellness product that accepts Illinois users falls within scope regardless of where the company is based, because the Act reaches the offering of services to people in the state. For national operators, that makes the safer engineering assumption a single high standard: keep a licensed professional in control of clinical decisions and therapeutic conversation everywhere, rather than geofencing compliance to one state. Illinois has also signaled a direction that other legislatures are watching, so the WOPR Act's structure is worth understanding even for providers with no current Illinois footprint.
The through-line is that the law does not ask whether an AI tool is helpful or accurate. It asks what the tool does. If the tool decides, converses therapeutically, plans treatment without licensed review, or reads emotion, it is on the wrong side of the line in Illinois.
Frequently Asked Questions
What did the Illinois WOPR Act actually change?
It made it unlawful to provide, advertise, or offer therapy or psychotherapy in Illinois unless a licensed professional delivers the service, and it barred AI from making independent therapeutic decisions, engaging in therapeutic communication with clients, generating treatment plans or recommendations without licensed review, and detecting emotions. AI is limited to administrative and supplementary support. The law is HB 1806, Public Act 104-0054, effective August 1, 2025.
Who is affected by the Act?
Licensed therapists and clinics, teletherapy and behavioral-health platforms, AI mental-health startups, and any company marketing therapy services to Illinois residents. Out-of-state platforms are covered if they offer services to people in Illinois.
Can a therapist use AI at all under this law?
Yes, for administrative and supplementary support such as scheduling, billing, reminders, and documentation assistance, subject to consent and disclosure conditions. What is prohibited is using AI to perform clinical work: independent therapeutic decisions, direct therapeutic communication with clients, treatment plans that reach a client without licensed review, and emotion or mental-state detection.
Is the WOPR Act the same as Illinois HB 3773 on AI in employment?
No. HB 3773 amends the Illinois Human Rights Act to address AI in employment decisions and runs on its own timeline. The WOPR Act, HB 1806, governs therapy and behavioral health. They are separate laws with different scope and should be handled as separate compliance tasks.
What are the penalties, and who enforces them?
The Illinois Department of Financial and Professional Regulation enforces the Act and may assess civil penalties of up to $10,000 per violation, based on the degree of harm and the circumstances. Because penalties are assessed per violation, platforms operating at scale face compounding exposure from a single noncompliant feature.
Sponsored Training
Browse the full AI Regulation News tracker
Informational analysis for working professionals, not legal advice. Confirm how any rule applies to your situation with qualified counsel.