Illinois Banned AI Therapy: What It Restricts and What Clinicians Can Still Do

What the Illinois AI therapy law does

The Illinois AI therapy ban restricts independent chatbot care and requires a licensed professional to deliver therapy services. On August 1, 2025, Governor JB Pritzker signed the Wellness and Oversight for Psychological Resources Act (HB 1806, Public Act 104-0054). It bans anyone from providing, advertising, or offering therapy or psychotherapy to the public in Illinois unless a licensed professional delivers it. The Illinois Department of Financial and Professional Regulation (IDFPR) enforces the law, with civil penalties up to $10,000 per violation. The core intent is consumer protection: the law draws a line around the licensed practice of therapy and says AI cannot stand in for a licensed professional.

For everyday concerns outside a crisis, a mental health assessment can help organize what you want to discuss with a professional. Grounding techniques offer a structured exercise without sharing personal information with a chatbot. These resources support reflection and skill practice; they cannot establish a diagnosis or determine immediate safety.

It generally aims to stop AI systems from independently providing therapy or psychotherapy services to the public, to prevent AI tools from being marketed or used as if they were a licensed therapist, to keep clinical decision-making, diagnosis, and treatment in the hands of licensed humans, and to protect consumers who may not realize they are talking to software rather than a trained professional.

The driving concern is that a vulnerable person seeking help should not end up relying on an unsupervised chatbot for clinical care that, by law and by training, belongs to a licensed clinician.

What the law does not do

It is easy to read headlines like Illinois bans AI therapy and assume every app just became illegal. That is not what happened. The restriction is narrow and aimed at AI acting as the therapist, not at every use of AI in mental health.

The law is generally understood to still allow AI that supports a licensed clinician: a therapist can use AI for scheduling, note-taking, or other behind-the-scenes tasks, as long as a licensed human remains responsible for the care. It also leaves room for general wellness and self-help tools, such as apps that offer journaling prompts, mood tracking, breathing exercises, or psychoeducation, which are different from a tool claiming to deliver therapy. Permission is limited to the statutory categories. Supplementary support excludes therapeutic communication, and using AI to record or transcribe a session requires written notice of the use and purpose plus the patient's or authorized representative's consent.

Personal use of AI chatbots is likewise not the target. The law regulates services offered to the public, not what an individual chooses to type into a general-purpose chatbot at home. The boundary that matters is between AI as a tool used by a professional and AI as the professional. Professional use is permitted only within the statute's defined limits. The second is what Illinois moved to restrict.

The act gives concrete examples: scheduling appointments and processing insurance claims are administrative support; organizing referrals and preparing client records are supplementary support. These permissions do not authorize direct therapeutic conversations with clients. The prohibition also covers AI detecting emotions or mental states, so a clinician should examine what a proposed feature actually does before adopting it. Source: the act's definitions and prohibited-use provisions.

Why states are restricting AI therapy

Illinois did not act alone. It joined other states that have started regulating AI in mental-health contexts. The pattern is driven by safety concerns.

The main concerns regulators point to include crisis handling, since a general chatbot is not built to recognize or respond safely to suicidal thinking, abuse, or acute risk the way a trained clinician can. Regulators are also reading what AI therapy studies show about crisis handling and risk. Concerns run highest around AI therapists for kids and teens. Another is accountability: licensed therapists answer to a board, carry ethical obligations, and can lose their license, while a chatbot has no license to lose. Regulators also worry about misleading marketing, where some tools blur the line and present themselves in ways that could lead a user to believe they are receiving real clinical care, and about privacy, since mental-health conversations are deeply personal and AI tools collect and process that data in ways many users do not fully understand.

Several states are passing similar rules, and the specifics differ. Nevada's AB 406 bars AI from providing therapy services, with fines up to $15,000 per violation, while Utah took the disclosure route, requiring mental-health chatbots to make clear that users are talking to AI. Some states focus on advertising claims. Some, like Illinois, focus on the practice of therapy itself.

What it means for consumers

If you live in Illinois, the practical effect is reassurance more than restriction. The law is meant to make sure that when you receive therapy, a licensed human is accountable for your care. Personal chatbot use for general help is outside the law's prohibition on providing unlicensed services. Whether a particular app is compliant depends on its functions and claims. What you cannot do is hire a chatbot as your licensed therapist, because no chatbot qualifies as one.

A few things are worth keeping in mind. AI tools can help with support between sessions, reflection, and learning, but they are not a replacement for professional care and not a crisis service. If you are in crisis or thinking about suicide, call or text 988; a chatbot is not built for that moment. Read how any app describes itself: if it implies it provides therapy or replaces a therapist, be skeptical. And check the privacy policy, because these tools collect sensitive information and you should know how it is stored and shared.

What it means for apps and developers

For companies building in this space, the message is to stay on the right side of the line. Tools positioned as wellness, coaching, journaling, or clinician-support products are in a different category than tools claiming to deliver therapy. The risk zone is marketing or operating a product as a standalone therapist for the public. A disclaimer does not override what a tool actually does. Describing a service as wellness support is insufficient if its functions meet the statute's definition of therapy or psychotherapy.

The practical steps many responsible apps already take track the emerging best practices for AI chatbots in therapy: being explicit that the tool is not a therapist or a substitute for professional care, making it obvious when a user is interacting with AI, and routing users toward human help and crisis resources when risk signals appear.

Penalties and definitions vary by state, so app makers in regulated states should get qualified legal guidance.

The bigger picture

The Illinois law reflects a question the whole field is wrestling with: how far should AI go in mental health, and where does it stop. The debate over AI therapy pros and cons is exactly what this law tries to referee. AI can help with access, affordability, and low-stakes support, but it cannot replicate the judgment, accountability, and human relationship at the center of real therapy.

Regulation like this aims to capture the benefits while protecting people from the risks. It tries to keep the door open to useful AI support while keeping clinical care in the hands of accountable, licensed humans.

What to ask before a clinician uses AI

Ask which task the software performs and whether it records or transcribes the appointment. If it does, ask for the written explanation of its purpose and how consent can be withdrawn. Also ask who checks the output and which information is stored. These questions distinguish appointment logistics from technology that processes the substance of your care.

The statute's permissions should not be read as authorization for a supervised chatbot to conduct an entire clinical conversation. Responsibility remains with the licensed professional, and the permitted support categories have defined limits. Consult the statute and IDFPR for the legal wording; a product's description of itself is not a compliance determination.

For general self-reflection, a mental health assessment can organize concerns for an appointment, and grounding techniques provide an exercise you can use without a chatbot. Neither establishes a diagnosis or replaces urgent help. The AI therapy guide covers the wider landscape, while the related state-bans and Nevada articles compare different legal approaches.

The act defines consent as an explicit, voluntary written agreement that can be revoked, including an electronic agreement. Accepting broad terms of use that mix an AI description with unrelated information does not meet that definition. Before agreeing, ask the practice to explain how you can revoke consent and what happens to recordings already created. The consent requirement and the separate restrictions on clinical use both apply. Source: the act's consent definition and permitted-use provisions.