AI Therapists for Kids and Teens: What Parents Need to Know for Safety

The short answer for parents

If you are wondering whether your child or teen should rely on an AI chatbot for mental-health care, involve a qualified human professional instead. Companion, roleplay, and general-purpose chatbots can produce unsafe advice and lack the clinical accountability of a licensed child therapist. APA's advisory on AI and adolescent well-being calls for protections suited to young people's development. A product's age rating or claim of being supportive does not demonstrate clinical safety for your child.

The reason is simple: a developing child or teen in distress needs a trained human who is accountable, who can recognize danger, and who can act. An AI cannot do any of that reliably. If your child is struggling, the safest response is to involve a licensed child or teen therapist, a school counselor, or a pediatrician, and to use a crisis line if there is any risk to their safety. If your child is in immediate danger or talking about suicide, call or text 988 in the US, or text HOME to 741741 for the Crisis Text Line. Call 911 for an emergency.

Parents can use the Stress Level Test to reflect on their own strain while arranging support; it should never be used to diagnose a child. Emotional Regulation Activities for Kids offers a way to practice naming feelings together. Choose activities your child can comfortably understand and stop if an exercise increases distress.

Why kids and teens turn to AI chatbots

Understanding why young people reach for these tools helps you respond without panic or blame. AI chatbots are free or cheap, available at three in the morning, can feel less judgmental, and can keep responding without a scheduled appointment. For a teen who feels alone, embarrassed, or afraid of being a burden, that can feel safer than telling a parent or a counselor.

Many teens also face real barriers to human help: long waitlists for therapists, cost, stigma, or simply not knowing how to ask. Companion and roleplay apps are designed to feel personal and to keep users engaged, so a lonely young person can quickly form an attachment. If your child has reached for one of these apps, the pull is strong and the tools are easy to reach, which is exactly why parental awareness matters more than blame.

The serious documented risks for minors

The risks here are well documented. Lawsuits allege that AI companion chatbots contributed to serious harm to teenagers who leaned on Character.AI as a therapist or companion, including a widely reported wrongful-death suit filed in 2024 after a 14-year-old died by suicide following extensive chatbot use. In January 2026, Character.AI and Google agreed in principle to settle that case and four related ones. In 2025, the US Federal Trade Commission opened an inquiry into several companies offering AI companion chatbots, with a specific focus on effects on children and teens. Together, these cases and investigations describe a pattern across multiple products and companies. A settlement or inquiry establishes neither a clinical harm rate nor proof that every allegation was adjudicated. The cases raise safety questions that families should take seriously while keeping allegations distinct from research findings.

The core safety failures fall into a few categories. First, chatbots can give harmful or dangerous advice, because they generate plausible-sounding text rather than clinically sound guidance, and they can affirm a young person's worst thoughts instead of challenging them. Second, they often fail to handle self-harm and suicidal crisis safely: a tool may miss the warning signs, respond inadequately, or continue an unsafe conversation instead of escalating to a human or a crisis line. Third, these systems are not built around child development or child-safety standards, so age-inappropriate, sexualized, or manipulative content can surface.

Privacy adds another layer of risk: children's deeply personal disclosures, including details about their mental health, may be stored, used to train models, or exposed, and minors cannot meaningfully consent to that. Dependency compounds it: companion bots are engineered to feel attentive and to maximize engagement, which can foster a parasocial bond, an unhealthy reliance, and withdrawal from real friendships and family at the very age when those human connections matter most.

Why AI is not an appropriate therapist for a child

A licensed child or teen therapist is trained, supervised, and professionally accountable. Their responsibilities include assessing risk, explaining confidentiality, and following applicable reporting and emergency requirements. They can consider tone, body language, family circumstances, and the things a child does not say, involve parents and schools when appropriate, and adjust care as a child develops. A chatbot cannot assume those professional duties. Vendor responsibilities and clinician reporting rules depend on applicable law; a reassuring response on a screen does not establish that anyone is monitoring or able to intervene.

Therapy for young people is also relational and developmental. It depends on trust built over time with a consistent, real person who understands the child's family, school, and stage of growth. A chatbot can imitate warmth, but it does not understand your child, cannot form a genuine therapeutic relationship, and is not a substitute for one. For these reasons, AI tools should be treated, at most, as general information, never as treatment for a minor. For adults, weighing AI therapy pros and cons can be a reasonable exercise; for a minor, it is not a close call.

Warning signs to watch for

You do not need to monitor every message to stay aware. Watch for patterns. Signs worth your attention include a child spending long stretches talking to an app late at night, becoming secretive or defensive about a chatbot or companion app, or describing the AI as a best friend, partner, or the only one who understands them.

Also notice withdrawal from real friendships and family, rising distress after using an app, or any mention of self-harm, hopelessness, or suicide whether to you or to a bot. If you see these signs, stay calm and open the conversation gently rather than confiscating the phone and shutting down. Ask what they get from the app and how they are really feeling. If there is any sign of self-harm or suicidal thinking, treat it as urgent and contact a professional right away, or reach out to 988 or the Crisis Text Line, both available 24/7.

Safer alternatives and how to get help

Better options exist and are within reach. A licensed child or adolescent therapist is the option with real training, supervision, and accountability; your pediatrician can refer you, and many therapists offer telehealth, which can shorten waits and ease cost. School counselors are free, already know your child's environment, and can be a fast first step. Our guide to AI therapists for schools examines proposals to use digital support in education and the questions families should ask. For lower-stakes support, well-vetted tools built around best practices for AI chatbots in therapy may have a limited role, but only alongside human care and with a parent involved, never as a replacement.

Keep crisis resources visible and easy to use. In the US, anyone can call or text 988 to reach the Suicide and Crisis Lifeline, 24 hours a day, or text HOME to 741741 to reach the Crisis Text Line. For an immediate emergency, call 911. The most protective thing you can offer is a steady, judgment-free relationship: let your child know they can come to you, and that getting real help is a sign of strength. If you would rather start with a professional, browse licensed child and teen therapists in our directory.

How to talk with your child about a chatbot

Start with curiosity: "What do you like about talking to it? How do you feel afterward? Has it ever said something that worried you?" Listen for the need underneath the app use, such as loneliness, embarrassment, or wanting someone available at bedtime. Agree on a trusted adult they can contact when a conversation becomes upsetting.

Review the app together when appropriate. Check what information it saves, whether conversations are used for training, how to delete the account, and whether safety settings can be changed. Explain that a reply about getting help does not mean a counselor has been contacted. Avoid promising complete secrecy if someone is in danger.

When arranging an appointment, describe changes in sleep, school participation, friendships, and mood, including whether chatbot use appears connected. NIMH recommends professional help when persistent difficulties interfere with everyday life. You do not need to prove that the app caused a problem before seeking support. If there is an immediate safety concern, use the crisis guidance above instead of waiting for an appointment.