Where people take their worries now
What new surveys say about AI, health and wellbeing, and the questions they leave open
Picture a teenager at the end of a long day.
The house is quiet. A worry has been circling for hours. It feels too big for a text to a friend and too small for a phone call to anyone official.
So they open an app and type it out.
That scene is not rare anymore. New surveys show how common it has become, and they raise a question worth sitting with: what does it mean when so many people bring their wellbeing to a machine?
What young people are doing
In June 2026, RAND researchers reported in JAMA Pediatrics on a nationally representative survey of 1,009 young Americans aged 12 to 21, fielded in November 2025.
The headline:
19.2% said they had used an AI chatbot for mental health advice. The authors compared this with 13.1% in a similar 2024 survey, while noting that the earlier survey used slightly different wording.
Among users, 42.8% said they did so at least monthly. And 63.3% said they had told no one.
Most users rated the advice as helpful: 91.7% said somewhat or very. But that is perceived helpfulness, not evidence of benefit. The authors warned that chatbot sycophancy, or overflattery, could influence those ratings. A kind answer is not always a right one.
So read the finding for what it is. It tells us what young people are doing, quietly, and often alone.
What adults are doing
The pattern is not limited to the young.
In a KFF poll published March 25, 2026, 32% of US adults said they had used AI for physical or mental health information or advice in the past year. About 29% had asked about physical health. 16% had asked about mental health. Among adults aged 18 to 29, 28% had asked about mental health or emotional wellbeing.
KFF also asked AI health users why. Among them, 19% said not being able to afford a provider was a major reason. And 18% said not having a regular provider, or not being able to get an appointment, was a major reason.
For some users, cost and access barriers were among the reasons they turned to AI.
These are numbers about behavior. They show where people turn. They do not show whether the turning helps.
Hope, and a note of caution
How do Americans feel about it?
Not warmly, it turns out. In an August 25, 2026 report, the Pew Research Center shared a survey of 3,488 US adults, fielded in late June. It asked whether chatbots do more to help or hurt people who use them for support with loneliness, depression or stress. These are public perceptions, not measured outcomes.
More respondents said hurt than help:
- Loneliness: 39% said hurt, 19% said help.
- Depression: 36% hurt, 17% help.
- Stress: 29% hurt, 22% help.
Between a quarter and a third said they were not sure.
Notice the shape of this. Many people are using a tool that many of their neighbors are wary of. Both things are true at once. Curiosity and doubt are moving through the same communities, often through the same households.
The question underneath the numbers
We do not know what AI will do for human wellbeing. The surveys above cannot tell us, and they do not claim to.
But they do point at something we can see clearly. People are looking for someone to talk to. They are looking for information they can trust. And some of them are looking in places the health system has not reached.
That is a human story before it is a technology story.
Which makes the real questions very human ones:
- Who is making sure young people have a trusted person to talk to, so the chatbot is not the only one listening?
- Who makes sure that when people reach for help, they can tell good information from agreeable information?
- Who decides what counts as evidence, and whose experience is counted?
Technology will keep changing. These questions will not.
The world we are working toward
Imagine a world where reaching for help always finds a human being on the other side. Where communities, families and young people are equipped to use new tools wisely, and to put them down when a person is what is needed. Where leaders ask what people actually need before deciding what to build.
That is the world the Global Wellbeing Institute is working toward. We believe wellbeing grows where communities are strong, people are resilient, and leaders are ready to listen.
If you lead a community, a school or a program, or you simply care about how people find support, we would love to hear from you. Come and think it through with us.
Get in touchdirector@globalwellbeinginstitute.org
Sources
- RAND study in JAMA Pediatrics (McBain, Cantor, Breslau and colleagues), published 1 June 2026: https://jamanetwork.com/journals/jamapediatrics/fullarticle/2849307
- RAND press release on the study, 1 June 2026: https://www.rand.org/news/press/2026/06/nearly-1-in-5-us-adolescents-and-young-adults-use-ai.html
- KFF Tracking Poll on Health Information and Trust: Use of AI for Health Information and Advice, published 25 March 2026 (fieldwork 24 February to 2 March 2026; 1,343 US adults): https://www.kff.org/public-opinion/kff-tracking-poll-on-health-information-and-trust-use-of-ai-for-health-information-and-advice/
- Pew Research Center, 25 August 2026 (survey 22 to 28 June 2026, 3,488 US adults): https://www.pewresearch.org/science/2026/08/25/do-americans-think-chatbots-help-or-hurt-people-using-them-for-loneliness-depression-or-stress/
Global Wellbeing Institute is a new organization and its programs are in development. We do not report program results.
