States Are Making AI 'Therapists' Illegal. Here's What the 2025 Laws Do.

In 2025, several states moved to make it illegal for an AI to act as your therapist. Illinois and Nevada passed laws barring AI systems from delivering mental health therapy to the public, and Utah now requires mental health chatbots to disclose that they are not human. The common thread is a simple line: an AI can help a clinic with scheduling, but it cannot be your licensed therapist.

Here is what the new laws do and what it means for you.

Know which states acted, and what they did

Three states led the way in 2025, with different approaches.

• Illinois passed the Wellness and Oversight for Psychological Resources Act (the WOPR Act), signed in August 2025. It bars providing therapy to the public unless a licensed professional delivers it, and stops clinicians from letting AI make therapeutic decisions or talk directly with clients as the therapist. AI can still handle administrative work like scheduling and notes.
• Nevada passed AB 406, effective July 2025, which prohibits AI systems from providing services that amount to the practice of mental or behavioral health care, and bars claiming an AI can.
• Utah took a lighter approach with HB 452, requiring mental health chatbots to clearly disclose that a user is talking to AI, and adding rules around using or selling personal health data.

In short, Illinois and Nevada effectively say an AI cannot be a therapist, while Utah focuses on honesty and privacy.

Understand why lawmakers drew the line

These laws did not come out of nowhere. They followed a wave of people using general chatbots for emotional support, and growing alarm from clinicians and regulators about the risks. The American Psychological Association warned in late 2025 that AI chatbots are not a substitute for professional mental health care.

The core concerns are consistent: an AI cannot reliably recognize a crisis or assess safety, it is built to be agreeable rather than to challenge a harmful thought, and no one is professionally accountable for your care the way a licensed clinician is. Lawmakers essentially wrote that reasoning into the rules.

See what the laws still allow

It is worth being precise, because these are not bans on technology in general. The Illinois and Nevada laws still let licensed practices use AI for administrative tasks, things like scheduling, billing, and paperwork, and let clinicians use AI tools under their own professional judgment and review.

What they prohibit is the AI standing in for the therapist: making treatment decisions on its own, or presenting itself to the public as something that can provide therapy. The relationship and the clinical judgment have to come from a licensed human.

Know what this means for you

Even if you do not live in Illinois, Nevada, or Utah, the takeaway carries over. These laws are not arbitrary; they reflect a broad professional consensus that an AI is not equipped to be your therapist, no matter how helpful it feels at 2am.

That does not mean a chatbot is useless. Used with clear eyes, it can work like a journal that talks back, helpful for sorting a thought or preparing for a real appointment. The trouble starts when it becomes your main place to process hard feelings, or your plan in a crisis, because that is exactly what it cannot safely be.

Find a real licensed therapist

If the appeal of a chatbot is that it is easy and immediate, the better answer is to make real therapy easier, not to settle for software. That is what TheraVoca is built to do. You describe what you need once, stay anonymous until you choose who to talk to, and your request goes to licensed Idaho therapists who fit and have openings, including many who work by video.

A real therapist gives you what the law is trying to protect: an accountable relationship, honest pushback, the ability to recognize a crisis, and actual treatment. You can get matched when you are ready, and what to expect from therapy shows how the early sessions tend to go.

Questions people ask

Did some states really ban AI therapy?
Yes. In 2025 Illinois (the WOPR Act) and Nevada (AB 406) passed laws barring AI systems from providing mental health therapy to the public, and Utah (HB 452) requires chatbots to disclose they are not human. They are among the first laws of their kind.

Does this mean I cannot use ChatGPT for mental health at all?
No. These laws target services that offer themselves as therapy, not your personal use of a chatbot as a journal or for general information. The point is that an AI is not a substitute for professional care, which the American Psychological Association also warned in late 2025.

Why are AI chatbots considered risky for therapy?
Because they cannot reliably assess safety or recognize a crisis, they tend to agree with you rather than challenge harmful thinking, and no one is professionally accountable for your care. Those limits are built into what a chatbot is.

What can clinics still use AI for?
Under the Illinois and Nevada laws, licensed practices can still use AI for administrative tasks like scheduling, billing, and notes, and clinicians can use AI tools under their own review. What is off limits is letting the AI act as the therapist.

What if I am in crisis right now?
An AI is the wrong tool for a crisis. If you are thinking about harming yourself, call or text 988, the Suicide and Crisis Lifeline, any time, day or night.

Let's recap

• In 2025, Illinois and Nevada passed laws barring AI from acting as a therapist, and Utah now requires chatbots to disclose they are not human.
• The laws reflect a clear concern: AI cannot reliably assess safety, tends to agree rather than challenge, and is not professionally accountable.
• They still allow AI for administrative work and as a clinician tool; what they bar is the AI standing in for a licensed therapist.
• The takeaway holds everywhere, and TheraVoca can match you to a real licensed Idaho therapist. In a crisis, call or text 988.