- In a nationally representative survey, 19.2% of adolescents and young adults ages 12 to 21 said that they had used AI chatbots for mental health advice in 2025.
- Most respondents (63.3%) said they had not discussed their AI chatbot use for mental health advice with anyone.
- Researchers recommended that mental health clinicians routinely ask their patients about AI chatbot use.
Roughly one in five young people used an artificial intelligence (AI) chatbot for mental health advice last year, a cross-sectional, nationally representative survey showed.
Among youths ages 12 to 21, 19.2% said that they had used AI chatbots for mental health advice in 2025, up from 13.1% in a similar 2024 survey, reported Ryan K. McBain, PhD, MPH, of the RAND Corporation in Arlington, Virginia, and co-authors in JAMA Pediatrics.
Among those who used AI chatbots for this purpose, 42.8% did so at least monthly, and 91.7% rated the advice as somewhat or very helpful. Most respondents (63.3%) said they had not discussed their AI chatbot use for mental health advice with anyone.
"The speed of growth is attention-grabbing, but so is the fact that most young people who use these tools for mental health advice say they are not telling anyone," McBain said in a press release.
This study "underscores the urgency of understanding and shaping the evolving role of AI chatbots in youth mental health care," the authors concluded. "As these technologies become increasingly integrated into the daily lives of young people, they should be understood as active contributors in the broader ecosystem of psychological interventions."
There are many perceived benefits to the use of chatbots for mental health, said co-author Hao Yu, PhD, of Harvard Medical School in Boston. Unlike human therapists, they have "endless patience" and are available in mental health professional shortage areas and accessible 24/7, he noted. However, "these tools can be dangerous," Yu told MedPage Today.
The fundamental problem is that chatbots are not developed as mental health tools, and are not designed to diagnose and treat health issues, and can provide both dis- and misinformation, Yu said.
Also, the AI landscape, which is rapidly changing, has few, if any, guardrails, he noted. Last year, the parents of 16-year-old Adam Raine sued OpenAI and its CEO, alleging that ChatGPT guided the boy in taking his own life.
Teenagers can ask chatbots anything, and if they show signs of suicidality, chatbots may encourage suicidal behavior instead of warning against it, Yu said.
As for the study's implications, "I would expect a mental health clinician now to routinely ask their patients, 'Have you ever used AI chatbots for mental health issues?'" he added.
The survey showed that use of AI chatbots for mental health advice was more common among females than males (adjusted OR 2.10, 95% CI 1.36-3.23), among those ages 18 to 21 versus those ages 12 to 14 (aOR 3.65, 95% CI 1.98-6.74), and among those who had spoken with a physician about their mental health in the prior 6 months compared with those who had not (aOR 1.89, 95% CI 1.18-3.03).
Young adults' more prevalent use of chatbots may be due to a higher prevalence of mental health conditions. It may also be a function of their greater access to smartphones and reduced parental supervision as children age, McBain and team noted.
Interestingly, young people who reported using AI chatbots for mental health support at least once per month were five times more likely to be Black (aOR 5.45, 95% CI 1.44-20.66) than white. Black youth may believe that "professionals are not as responsive to their unique needs," the authors speculated. Alternatively, the greater reliance on chatbots in this population might also signal "reduced access to professional services."
For this study, the authors drew participants from members of RAND's American Life Panel. The survey was conducted among adolescents and young adults ages 12 to 21 in November 2025. They included a U.S. population-weighted 42,825,655 youth (unweighted 1,009 youth). Median age was 17 years, 50% were male, 49.8% were white, 25.7% were Hispanic, and 13% were Black.
The study had a relatively small sample size and a completion rate of 58.4%, suggesting that the findings "contain uncertainty and may be subject to nonresponse bias," the authors noted.
McBain said the survey did not investigate the quality of care that chatbots delivered, which could be an area of future study. Yu noted that another future research question is why certain populations used chatbots more than others.
Disclosures
This work was supported by grants from the National Institute of Mental Health, including to McBain.
McBain reported no other disclosures. Co-authors reported relationships with the National Institute of Mental Health, the National Institute on Aging, Chestnut Health, the Government Accountability Office, the Aspen Institute, the NYS Youth Justice Institute, Pfizer, GSK, Black Opal, the Peterson Health Technology Institute, the National Institute on Minority Health and Health Disparities, the National Institute on Alcohol Abuse and Alcoholism, and the National Institute of Nursing Research.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.