Nearly 1 in 5 Young People Have Used AI Chatbots for Mental Health Advice

AI chatbots are becoming part of how some teenagers and young adults deal with emotional stress. A national study published in JAMA Pediatrics found that 19.2% of Americans ages 12 to 21 had used an AI chatbot for mental health advice as of 2025, representing an estimated 8.2 million young people. Among those who had used one, 42.8% did so at least monthly, while 91.7% described the advice as somewhat or very helpful. The appeal is understandable: chatbots are available around the clock, respond immediately, and may feel easier to approach than another person when someone is embarrassed, anxious, or simply wants privacy. But researchers stressed that widespread use does not establish that chatbot advice actually improves mental health outcomes.

The same study found that 63.3% of users had not told anyone they were using AI for mental health advice. That matters because parents, doctors, counselors, or other trusted adults may be unaware of information influencing a young person’s decisions. The survey also found that youths who had recently discussed their mental health with a physician were more likely to use chatbots, suggesting AI may sometimes supplement rather than replace conventional support. Still, the study measured self-reported use and perceived helpfulness—not whether the advice was clinically accurate, safe, or effective. Researchers therefore called for parents and healthcare professionals to ask about chatbot use directly and help young people understand its limitations.

Chatbots can provide general information, help someone organize thoughts, or offer another perspective, but they can also give incorrect information, reinforce reassurance-seeking or unhealthy beliefs, and miss important signs of serious risk. Adolescents may also be more likely than adults to interpret simulated empathy as genuine understanding or place too much trust in an AI response. A safer approach is to treat AI as a limited supplementary tool while keeping human support—from family, trusted adults, doctors, counselors, or therapists—within reach, especially when distress is persistent, worsening, or involves thoughts of self-harm.

If a young person in your life uses AI for emotional support, make the conversation about safety and trust—not punishment—and remind them that a chatbot should never be their only source of help.