AI is shaping health and human development
We study what that means for children, families, and adults — and who is affected most.
A research collaboration at Brigham and Women’s Hospital, Harvard Medical School, and Fordham University.
We study what that means for children, families, and adults — and who is affected most.
A research collaboration at Brigham and Women’s Hospital, Harvard Medical School, and Fordham University.
Our work
Artificial intelligence is entering how people learn, seek information, and manage their health — for children encountering it earlier than any generation before them, for the families around them, and for adults navigating it in their own lives.
AI is becoming a new kind of context for human development: a personalized, responsive presence that people talk to, learn from, and lean on. It responds like a relational partner, and people form real relationships with it and draw on it in shaping who they are.
How children, adolescents, and young adults engage with generative AI and what it means for their mental health — who turns to it when lonely, and whether it replaces or supplements human relationships.
How adults and young adults decide whether to trust and rely on AI for health information and support, drawing on a nationally representative sample of more than 10,000 U.S. adults and separate studies of college students.
In our survey, Asian students had about twice the odds of using AI for mental health. They remain among the least studied. We center their experiences.
Our framework
Generative AI differs from earlier technologies in producing reciprocal, personalized responses that can function as a relational partner. Our framework in JAMA Pediatrics sets out how that intersects with four core developmental tasks of adolescence.
Who young people turn to, and what it means when something always available becomes one of those sources.
How distress gets managed, and whether a responsive system supports coping or substitutes for it.
How young people read intent, perspective, and reciprocity in something that responds like a person.
How self-understanding develops when a system reflects a version of you back.
Liu, C. H., & Yip, T. (2026). Generative AI in adolescence: A developmental framework. JAMA Pediatrics, 180(5), 473–474. Descriptions are our summaries — check against the paper before launch.
What we found
And the students who are struggling most are the most likely to. In our survey of 675 students, moderate depression, severe depression, severe anxiety, and suicidality were each associated with roughly twice the likelihood of turning to AI for mental health.
Liu, C. H., Zhang, W., Lou, F., Zhao, C., Chow, A., & Yip, T. (2026). Clinical and sociodemographic predictors of AI use for mental health among college students. Journal of Affective Disorders.
“College students who are most drawn to AI for mental health may also be the most vulnerable to its risks.”
Cindy H. Liu, PhD · Mass General Brigham
“If they don’t know how to have human relationships … with the friction and the disagreement and the not-constant availability, that could have implications for how they develop relationships in the real world.”
Tiffany Yip, PhD · Fordham Now
What we are finding
In a nationally representative study of 10,506 U.S. adults, most people trusted AI less than a human professional across every health domain we asked about. But the gap was widest for mental health counseling: fewer than one in eight would trust AI as much as a person, against more than four in ten for diet and exercise planning.
Share who would trust AI as much as, or more than, a human professional. Liu, Koire & Yip, JMIR (invited revision under review).
A pattern across datasets
In a national sample of adults and in two separate college student samples, Asian respondents are consistently more likely to trust AI for health and to turn to it for mental health and companionship. The estimates come from different populations, different measures, and different studies, and they point the same way.
Why is the open question. Stigma around disclosure, concern about burdening others, and difficulty finding culturally and linguistically matched care are all plausible. AI asks nothing, judges nothing, and costs nothing to try. Whether that fills a gap in care or leaves people outside it is what we are working to find out.
Estimates for Asian respondents, adjusted, from three independent datasets. Liu et al. (2026), JMIR; Journal of Affective Disorders; Computers in Human Behavior: Artificial Humans.
Who this is for
We are clinical and developmental psychologists and researchers. What we can do is make sure the questions we ask come from the people living them, and that the answers get back to families, clinicians, and schools in a form they can actually use.
We have a strong record of advisory work in youth mental health, including chairing the advisory board of a national youth mental health AI tracker. We intend to convene a youth advisory board for this initiative.
Alongside our work in scholarly journals, our framework appeared in JAMA Pediatrics and Pediatrics — reaching the pediatricians and mental health clinicians families actually talk to.
When findings matter to parents and young people, we say so publicly, in plain language, rather than leaving it inside a paywall.
In our survey, Asian students had about twice the odds of using AI for mental health, yet they remain among the least studied. We are intentional in building evidence with Asian Americans in view.
Latest
“Study Finds Students with Highest Distress Use AI for Mental Health at Elevated Rates.” Mass General Brigham.
Read →“Students in Distress Turning to AI for Mental Health Support, Study Shows.” Fordham Now.
Read →“Study Finds Students with Highest Distress Use AI for Mental Health at Elevated Rates.” Harvard Brain Science Initiative.
Read →“The Robot Friend? Study Warns Teens Increasingly Use AI Chatbots for Emotional Support.” NBC Palm Springs.
Read →“College students’ trust in generative AI for mental health.” Psychiatric Services.
“AI and teen identity formation.” Pediatrics. Embargoed until Aug 24.
Tiffany Yip on students’ use of AI for mental health, live on SiriusXM Doctor Radio’s “Psychiatry” with Dr. Thea Gallagher, 11 a.m. ET.
Journal of Research on Adolescence special issue, “Youth Development in the Age of Generative AI.” Guest edited by Angela Chow, Cindy Liu, and Thao Ha. Empirical submissions only.
Submit an abstract →Who we are
Director of the Developmental Risk and Cultural Resilience Lab at Brigham and Women’s Hospital. A clinical psychologist whose work examines how mental health takes shape within relationships — family, cultural, and now technological — from the perinatal period through young adulthood, supported by the National Institutes of Health. Clarivate Highly Cited Researcher, elected Fellow of the Academy of Behavioral Medicine Research, and recipient of the Harvard Medical School A. Clifford Barger Excellence in Mentoring Award.
Director of the Youth Development in Diverse Contexts Lab. A community developmental psychologist whose work examines identity, culture, and the daily and physiological processes that shape adolescent and young adult development, supported by the National Institutes of Health and the National Science Foundation. Elected Fellow of the Academy of Behavioral Medicine Research and the Association for Psychological Science, and recipient of the American Psychological Association’s Distinguished Career award.
Selected publications
Get involved
We welcome researchers across disciplines, the people building and studying these systems, schools and clinicians, and funders who want young people and families represented in the evidence that will shape how AI is used.
Researchers, clinicians, schools, and teams building these systems.
Start a conversation →Philanthropic support is arranged through our institutions. We are glad to talk about what it would fund.
Talk to us →Cindy H. Liu · [email protected] Tiffany Yip · [email protected]