Quantifying Generative AI Use for Mental Health Advice Among Adolescents

New research surveyed adolescents and young adults in the US on their use of generative artificial intelligence for mental health advice when experiencing emotional distress. Of individuals surveyed, 13% reported using AI for mental health advice, with ages 18 to 21 showing higher rates of use.

The survey assessed whether respondents had used generative AI generally, if they had asked for advice or help from AI when feeling sad, angry, or nervous, how often they sought such advice, and how helpful they perceived the advice to be. The survey provided a definition of generative AI and examples of common AI platforms. After collecting responses, investigators calculated survey-weighted percentages and conducted multivariable logistic regression to look at key factors associated with AI use for mental health advice. Factors included were age, sex, race and ethnicity, highest parental education level, parental marital status, and census region. Questions used plain language terms for emotional states to ensure comprehension in the entire age group.

The population surveyed was adolescents and young adults aged 12 to 21. Of 2125 individuals contacted for the survey, 1058 responses were recorded. Respondent age demographics were 37% aged 18 to 21, 33% aged 15 to 17, and 30% aged 12 to 14. Gender and racial demographic were 50.3% female, 13% Black, 25.2% Hispanic, and 51.3% White. Participants were taken from the RAND American Life Panel and Ipsos Knowledge Panel, 2 nationally representative survey panels in the US; individuals were randomly chosen from these groups to be reached out to for participation in the survey. The study did not include any measurements of diagnosed mental illness, and investigators noted that findings are only generalizable to English speakers aged 12 to 21 with internet access.

Investigators found 13% of respondents reported using generative AI to seek mental health advice. Among ages 18 to 21, 22.2% of respondents reported use of AI for mental health advice. For those who reported this use, 65.5% sought advice monthly or more often. In terms of usefulness of advice, 92.7% of respondents reported they found the AI advice somewhat helpful or very helpful. Reporting the advice was somewhat helpful was the most common response across all age subgroups. Generative AI use for mental health advice was higher among ages 18 to 21 (P < 0.001) compared with younger groups. Black participants were less likely to rate the AI advice as helpful (P = .01) compared with White non-Hispanic participants. There were no other statistically significant group differences.

Investigators noted in this new paper that “high use rates likely reflect the low cost, immediacy, and perceived privacy of AI-based advice, particularly for youths unlikely to receive traditional counseling. However, engagement with generative AI raises concerns especially for users with intensive clinical needs, given difficulties in establishing and using standardized benchmarks for evaluating AI-generated mental health advice and limited transparency about the datasets training these models.”^1,2,3

References

1. McBain RK, Bozick R, Diliberti M, et al. Use of generative AI for mental health advice among US adolescents and young adults. JAMA Netw Open. 2025;8(11):e2542281.

2. Abbasian M, Khatibi E, Azimi I, et al. Foundation metrics for evaluating effectiveness of healthcare conversations powered by generative AI. NPJ Digit Med. 2024;7(1):82.

3. Lawrence HR, Schneider RA, Rubin SB, et al. The opportunities and risks of large language models in mental health. JMIR Ment Health. 2024;11:e59479.