AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults.
Level 4 - case-series / case-control
Nationally representative cross-sectional survey
PubMed 42223976 · doi:10.1001/jamapediatrics.2026.2015
What was done
A cross-sectional, nationally representative survey was conducted in November 2025 among 1,009 US adolescents and young adults aged 12 to 21 years (population-weighted to 42,825,655 youth; median age 17). The study assessed self-reported prevalence, frequency, perceived helpfulness, and disclosure of AI chatbot use for mental health advice. Multivariable logistic regression was used to analyze demographic and geographic correlates, as well as prior physician discussions about mental health.
What was found
An estimated 19.2% of respondents (weighted n = 8,207,180) reported using AI chatbots for mental health advice. Among users, 42.8% sought advice at least monthly, 91.7% rated it as somewhat or very helpful, and 63.3% had not disclosed this use to anyone. Use was independently associated with female sex compared with male sex (aOR 2.10; 95% CI, 1.36–3.23), age 18 to 21 years compared with 12 to 14 years (aOR 3.65; 95% CI, 1.98–6.74), and having spoken with a physician about mental health in the prior 6 months (aOR 1.89; 95% CI, 1.18–3.03).
Why it matters
Nearly a fifth of US youth turn to AI chatbots for mental health advice, the majority without disclosing it, underscoring the need for clinicians and families to proactively address AI use.
Limits
The study is cross-sectional and relies entirely on self-reported survey data, introducing potential recall and social desirability biases. The unweighted sample size was 1,009. The abstract does not evaluate the clinical safety, quality, or accuracy of the AI advice received, specific chatbot platforms, or objective clinical outcomes.