Rutgers psychiatrist warns AI reassurance can backfire for young mental-health users

A reassuring chatbot response is not always a helpful one. In a Rutgers Q&A published November 5, 2026, on Newswise, child and adolescent psychiatrist Hannah Simon warned that AI’s agreeable replies can reinforce unhealthy patterns. She also described a constructive role: helping young people understand their feelings and prepare to seek human support.

An open door when care is hard to reach

Simon, medical director of Rutgers’ Brandt Behavioral Health Treatment Center and Retreat, sees several reasons young people turn to AI. Chatbots offer immediate, around-the-clock information at little or no cost. They can also feel like a private, nonjudgmental place to ask questions that seem difficult to raise with parents or professionals.

That accessibility contrasts with the obstacles to psychiatric care. Simon cited shortages of mental-health professionals, especially child and adolescent psychiatrists, alongside lengthy waitlists, limited insurance coverage and substantial out-of-pocket costs. For young people facing those barriers, an immediately available answer has an appeal even though it cannot replace professional evaluation or treatment.

The Rutgers article cites a RAND survey estimating that about 8.2 million Americans ages 12 to 21—nearly one in five—use AI for mental-health information. Simon’s discussion focuses on how to handle that use, rather than presenting a new study of whether chatbot advice improves mental health.

When reassurance works against treatment

Simon’s chief concern is that chatbots can deliver inaccurate or misleading information with convincing confidence. A young person describing worry and trouble sleeping might receive a suggested psychiatric diagnosis. But overlapping symptoms require a fuller assessment, including medical issues and environmental factors, that a chatbot lacks the clinical judgment to provide.

Accuracy is not the only issue. In treating obsessive-compulsive disorder, Simon explained, clinicians often help patients tolerate uncertainty instead of repeatedly seeking reassurance. A chatbot can do the reverse: supply endless comfort that brings temporary relief while reinforcing anxiety and compulsive behavior. The reply may feel supportive while working against a treatment goal.

She also warned that reliance on a personal chatbot for emotional support can crowd out meaningful relationships and chances to develop independent coping skills. Used differently, however, AI can help someone find words for an experience they have struggled to explain, making a conversation with another person or a professional easier.

I often speak to patients about using AI as an educational resource rather than a medical authority.

Hannah Simon, Rutgers child and adolescent psychiatrist, in the Newswise Q&A

A boundary for users—and a conversation for parents

Simon sees educational value in learning about common symptoms, psychiatric terms, treatment options and coping strategies. Chatbots may also help users recognize when professional support is appropriate. Her guidance draws clear limits around that role:

• Do not rely on AI for self-diagnosis, medication decisions or managing a psychiatric crisis.
• Check important information with a qualified professional or established sources such as the National Institute of Mental Health.
• Be cautious about sharing sensitive personal information, because chatbot conversations may not be fully protected.

For parents, Simon recommends curiosity rather than punishment. Ask what a child is learning and whether the conversations help. She said some patients worry their parents do not believe in therapy, will judge them for needing treatment or cannot afford care. Turning to AI may signal that a young person recognizes a struggle and is looking for help—not necessarily a cause for alarm.

The warning signs she identifies concern what chatbot use does to daily life: disrupted sleep, schoolwork or socializing, and excessive worry about diagnoses. Her recommendation is to understand the need AI is meeting and help the young person reach appropriate support, rather than make stopping chatbot use the automatic goal.