About 1 in 5 adolescents and young adults (ages 12–21) in the U.S. reported using AI chatbots for mental health advice in 2025 — up from 1 in 8 just a year earlier. The trend is growing fast, but what's especially notable is how quietly: nearly two-thirds of teens who use these tools haven't told anyone, including parents, friends, or clinicians. A separate Canadian cross-sectional study of nearly 40,000 students found that affective AI use was independently linked to emotional distress and was nearly twice as likely to be associated with clinical-level emotional problems.

Mental health experts warn that when chatbots become a substitute for trusted adults or peers, they can create a problematic echo chamber — one that simulates empathy but lacks genuine understanding. AI tools are also designed to agree with users rather than challenge them, and the personal data shared with these platforms is stored and used to train commercial models, raising serious privacy concerns.

Key Takeaways:

• Clinicians should routinely and nonjudgmentally ask adolescent patients about AI chatbot use during visits — before taking a traditional history, to avoid anchoring on chatbot-generated "diagnoses"
• AI use for emotional support was more common among females, gender-diverse youth, Black and Indigenous students, and older teens
• Chatbot use should raise concern when it replaces — rather than supplements — human support systems

Why it matters: AI chatbots are already embedded in how teens navigate emotional distress, often invisibly. Without clinician awareness, these tools could delay appropriate care, reinforce harmful thinking patterns, or expose vulnerable youth to privacy risks — making proactive screening a critical new frontier in adolescent mental health.