Illinois has banned the use of AI in mental health therapy, making it the latest state to draw a hard line on chatbots in clinical care. The new law prohibits AI from being used to diagnose, treat, or communicate with patients in therapy sessions.
The Washington Post says Illinois is now aligned with Nevada and Utah, both of which have restricted AI therapy tools amid mounting safety concerns. The law blocks chatbots from being marketed as substitutes for licensed therapists.
Prioritizing public safety over rapid AI rollout
Under the measure, companies are prohibited from offering or advertising AI-powered therapy services unless a licensed professional is directly involved. The law bans the use of chatbots to diagnose, treat, or communicate with patients.
Violations will be investigated through the state’s complaint system and can bring civil fines of up to $10,000. Enforcement will be led by Mario Treto Jr., secretary of the Illinois Department of Financial and Professional Regulation, who told The Washington Post that his office will apply “the letter of the law” while “centering public safety first” and “balancing thoughtful regulation without stifling innovation.”
The statute does not outlaw all uses of artificial intelligence in mental health care. Licensed providers may still employ AI for administrative or non-clinical tasks.
How the law will be enforced is less straightforward. Complaints will trigger investigations, but the law leaves open questions about where the boundaries of therapy begin and end.
Treto has declined to outline specific scenarios, leaving companies and regulators to navigate a gray area as the rules take effect.
AI isn’t ready for therapy
More states are beginning to step in. In California, a proposal would set up a task force on AI and mental health, while lawmakers in New Jersey want to prohibit developers from promoting chatbots as licensed professionals.
The conversation comes against a backdrop of studies that show both potential and risk. In one trial, Dartmouth’s Therabot chatbot helped ease symptoms of depression and anxiety at rates comparable to traditional outpatient care. Talking to some AI tools, like Claude, also makes some users feel more positive.
Yet separate research from OpenAI and MIT Media Lab warned that prolonged reliance on voice-based bots, especially those designed with neutral tones, could deepen loneliness and foster unhealthy emotional attachment.
Research from King’s College London warns that AI chatbots can validate or even intensify delusional thinking, with reports of users developing psychotic symptoms after extended interactions. For regulators, such risks reinforce why AI, despite early promise, isn’t ready to serve as a therapist.
Bans are only half the answer
State bans are an important step, but the work cannot stop at legislation. AI companies must also do their part to keep users safe.
Anthropic is working with crisis support group ThroughLine to shape how its Claude chatbot responds to sensitive conversations, directing users toward real-world help when needed. The company says the tool is not a therapist, and safeguards are being built to keep it that way.
OpenAI has added mental health guardrails to ChatGPT, including reminders to take breaks, revised responses to emotional questions, and prompts that steer users toward reflection instead of direct advice.
At stake is a truth that cannot be overlooked: Laws may draw the boundaries, but safety will be decided in the code.