Should you turn to AI for mental health advice?

Lokesh Shahani, MD, PhD, psychiatrist and chief medical officer, UTHealth Houston School of Behavioral Health Sciences. (Photo by UTHealth Houston)

Welcome to “Ask the Expert,” a UTHealth Houston news series where our leading experts examine pressing health challenges. In this edition, we address using AI for mental health advice. 

The on-demand availability of artificial intelligence chatbots means that most individuals have quick, easy access to a wealth of information. The ease and accessibility of these tools may make it tempting to turn to AI for advice about mental health. But it’s important to note that these systems are prone to reinforcing a user’s existing beliefs or, in some cases, providing inaccurate information. 

AI chatbots should never be used to diagnose or treat mental health conditions. Instead, AI tools should be used as a supplement to professional care. 

What should adults know before turning to AI for mental health advice?

Before turning to AI for mental health advice, there are several things to keep in mind. AI can be confidently wrong. A chatbot can produce a compassionate-sounding answer that is inaccurate, incomplete, or inappropriate for a particular person. It does not have the clinical judgment of a licensed mental health professional. An AI also cannot truly assess the whole person. A clinician can consider history, behavior, nonverbal cues, medical factors, environment, and risk, while a chatbot generally sees only what the user types or otherwise provides.

A few other points of caution: 

• Don’t use AI to diagnose yourself. AI can help someone learn about possible symptoms or prepare questions for a professional, but an AI-generated diagnosis should not be considered reliable.
• Be careful with personal information. Mental health conversations can contain extremely sensitive information. Users should understand how a particular service stores, uses, and potentially shares their data before disclosing it.
• Watch for excessive dependence. Because chatbots are available 24/7 and can seem endlessly validating, someone can gradually begin relying on a chatbot instead of friends, family, or professionals.
• If AI gives advice about medication, symptoms, treatment, or a serious mental health concern, check it with a qualified health professional.

What should children and parents consider?

Children and adolescents may be less likely than adults to recognize that an AI’s apparent empathy isn’t genuine understanding or to question whether its advice is accurate. 

Parents should:

• Talk openly about AI rather than simply banning it. Ask what the child is using it for and what kinds of conversations they’re having.
• Teach children that AI can be wrong. A confident or friendly response isn’t evidence that it’s true.
• Set boundaries around AI companions and mental health chatbots. The concern isn’t just bad information; children can develop strong emotional attachments to systems that simulate friendship or intimacy.
• Protect your children’s privacy. Children may disclose things to a chatbot that they would never tell an adult. Parents should help them understand that an AI conversation isn’t necessarily private or confidential.
• Watch for behavioral changes. Warning signs can include withdrawing from friends and family, losing sleep because of chatbot use, abandoning normal activities, or treating the chatbot as a primary friend or authority.
• Keep human relationships central. AI shouldn’t replace parents, teachers, school counselors, pediatricians, friends, or other trusted adults.

Is there ever an appropriate time to ask AI for mental health advice?

If “advice” is understood as low-stakes support rather than diagnosis or treatment, an AI chatbot can be helpful. Reasonable uses might include:

• Asking for general information about a mental health topic.
• Brainstorming questions to ask a therapist or doctor.
• Organizing thoughts before an appointment.
• Journaling or reflecting on emotions.
• Learning general coping strategies and then deciding whether they are appropriate with a professional.
• Practicing a conversation—for example, how to tell a doctor or family member, “I've been feeling depressed.”
• Getting suggestions for healthy habits such as sleep, exercise, relaxation, or stress management techniques.

There are situations in which it’s unwise to turn to an AI bot for mental health advice. Don’t rely on AI to handle suicidal thoughts, self-harm, psychosis, threats of violence, severe eating disorder symptoms, or another mental health emergency. Those situations require human intervention.

Are some groups especially vulnerable to relying on AI?

Turning to AI for mental health advice is unwise for certain vulnerable populations. This includes children and adolescents because of their developmental vulnerability and the possibility of greater trust in AI-generated advice; people experiencing severe mental illness or disordered thinking, because an agreeable chatbot may inadvertently reinforce distorted or delusional beliefs; and socially isolated or lonely people, who may be particularly susceptible to replacing human relationships with an always-available AI companion. 

Certain mental health conditions, like anxiety or obsessive-compulsive disorder, may make some individuals more likely to experience negative outcomes if they rely on AI for mental health advice. AI bots, for example, can potentially reinforce reassurance-seeking, rumination, or compulsive behaviors.

People with limited access to mental health care, including people who are uninsured, underinsured, rural, or facing stigma, should also be cautious about relying on AI for mental health advice. Ironically, these people may benefit substantially from accessible information while also being at greater risk of having AI become their only source of support.

What other options are available besides AI?

Where an individual turns to for help depends on individual circumstances. Alternative options to AI include: 

• A trusted person: A friend, family member, teacher, clergy member, mentor, or other person you feel safe talking to.
• A primary care doctor: They can evaluate symptoms, rule out medical contributors, discuss treatment options, and refer you to mental health care.
• A licensed therapist, psychologist, counselor, or psychiatrist: These professionals can provide individualized assessment and treatment.
• School or university counseling services: Often a good first step for students.
• Community mental health centers and nonprofit organizations: Particularly useful for people concerned about cost or insurance.
• Peer-support groups: These can provide connection with people who have lived experience, although they aren’t a replacement for professional treatment when it is needed.
• Evidence-based mental health organizations: Organizations such as the American Psychological Association, National Alliance on Mental Illness, and the Substance Abuse and Mental Health Services Administration provide educational information and connections to resources.
• Crisis services: People in the U.S. experiencing a mental health crisis can call or text 988 for immediate connection with the 988 Suicide and Crisis Lifeline. For imminent danger or a medical emergency, call 911 or go to an emergency department.

The bottom line

AI can be a useful mental health tool, but it should not become a person's sole source of mental health care. Its availability, speed, and lack of perceived judgment are genuine advantages—but those same characteristics can make it especially easy to trust, overuse, or become dependent on. The safest approach is to keep humans in the loop, verify consequential information, protect sensitive data, and treat AI as a supplement rather than a substitute for qualified care.

By Lokesh Shahani, MD, PhD, psychiatrist and chief medical officer, UTHealth Houston School of Behavioral Health Sciences. All quotes should be attributed to him. 

For media inquiries or to suggest future health topics, contact UTHealth Houston Media Relations at [email protected] or 713-500-3030.