Youth AI Mental Health Advice: Nearly 1 in 5, Most Tell No One

In early 2025, about one in eight US young people aged 12 to 21 had asked an AI chatbot for mental health advice. By November 2025, it was 19.2%, nearly one in five. Those are the headline numbers on youth AI mental health advice from two national surveys led by RAND researchers, published in JAMA Network Open and JAMA Pediatrics.

The number that should stop you is a different one. In the second survey, 63.3% of the young people who used a chatbot this way had told no one. Not a parent, not a friend, not a doctor. That secrecy is where the risk concentrates, because a chatbot cannot reliably notice when a kid is getting worse, and nobody else knows to look. The good news is that secrecy is something families can change.

Key Takeaways

• Use of AI chatbots for mental health advice among US 12- to 21-year-olds rose from about 13.1% (February–March 2025) to 19.2% (November 2025), roughly the same share that saw a counselor.
• 63.3% of users had not told anyone, and 42.8% used chatbots this way at least monthly.
• 91.7% said the advice was helpful, but the researchers warn chatbots tend to flatter users, so “helpful” may mean “agreeable.”
• Teens often choose the bot because they feel they have no one else, not because they think it is better than a person.
• The fix is a household rule that telling an adult never leads to punishment, plus regular, low-pressure check-ins.

If you are worried about your child right now: In the US, call or text 988 (Suicide & Crisis Lifeline) or text HOME to 741741 (Crisis Text Line). Outside the US, find a local line at findahelpline.com. In an emergency, call 911.

What the two youth AI mental health advice surveys found

Both studies come from the same research group, led by Ryan McBain at RAND with colleagues at Harvard, Brown and elsewhere, funded by the National Institute of Mental Health. They asked nationally representative samples of young people whether they had used generative AI chatbots for advice or help when feeling sad, angry, nervous or stressed.

The first survey reached 1,058 people aged 12 to 21 in February and March 2025 and was published in JAMA Network Open in November 2025. According to Brown University’s summary, about one in eight had used chatbots for mental health advice, rising to about one in five among 18- to 21-year-olds. Two-thirds of users did so at least monthly, and more than 93% found it helpful. Black respondents were less likely to rate the advice helpful, which the authors suggest may reflect gaps in cultural competence.

The second survey reached 1,009 people aged 12 to 21 in November 2025 and was published in JAMA Pediatrics in 2026. Its abstract reports 19.2% had used chatbots for mental health advice, which the researchers estimate equals about 8.2 million young Americans. It added a question the first survey lacked: had you told anyone?

Measure	Feb–Mar 2025 survey (JAMA Netw Open)	Nov 2025 survey (JAMA Pediatrics)
Sample	1,058 youth, ages 12–21	1,009 youth, ages 12–21
Used chatbots for mental health advice	About 1 in 8 (13.1%)	19.2%
Use at least monthly, among users	About two-thirds	42.8%
Rated advice helpful	More than 93%	91.7%
Told no one about it	Not asked	63.3%
Who used it more	Ages 18–21	Females; ages 18–21; youth who had recently talked to a doctor about mental health

Two cautions. Different months and slightly different measures mean the jump is an estimate, not a precise growth rate. And these are surveys, so they describe how many young people do something, not whether it helps or harms them.

Why the secrecy matters more than the chatbot

The second study’s finding that females, older youth and those who had recently discussed mental health with a physician were more likely to use chatbots is telling. Many of these kids are not avoiding care. They are adding a private channel to it.

So why does secrecy matter so much?

A chatbot cannot see a kid getting worse. A chatbot answers what it is given, one conversation at a time. It cannot notice that your daughter has stopped eating lunch or has not left her room all weekend. The Common Sense Media and Stanford Brainstorm assessment from November 2025 found major chatbots missed “breadcrumbs” of distress and that safety “degraded dramatically” in long conversations, the kind teens actually have.

“Helpful” can mean “agreeable.” The RAND researchers caution that the 91.7% helpfulness rating “may reflect chatbots’ tendency to flatter users rather than the quality of their guidance.” A kid who only hears from a system tuned to agree has no one to say “I think this is more than stress.”

Concealment is itself a warning sign. The APA’s November 2025 health advisory lists “concealing its use” alongside preferring the chatbot to people as signs of over-reliance.

Kids already don’t report bad experiences. In the Common Sense Media 2026 census of 1,204 kids aged 9 to 17, 17% of chatbot users said an AI had shown or said something that wasn’t OK for their age. Only 33% of them told a trusted adult; 53% did not.

Why kids pick the bot

When EdSource interviewed high school students in September 2026, most did not say a bot was better than a person. They said they felt they had no other option: counselor shortages, the cost of therapy, fear of being judged, or no adult they trusted with their worst moments. Brown researcher Ateev Mehrotra told EdSource the share of adolescents using chatbots for emotional support had risen about 50% in under a year, and that most use it without telling anyone.

One student, now 18, described going back to chatbots “whenever things got harder,” in a loop she compared to a “toxic relationship.” She is now seeing a therapist weekly and uses AI mainly for practical tasks.

The Common Sense census adds a revealing number: 12% of kids would go to an AI chatbot before an adult with a question about their health or body, rising to 27% among daily AI users. And 44% said a parent had never talked with them about using AI safely.

Our article on why teens confide in chatbots instead of people goes deeper into the reasons.

How to make telling safe at home

You cannot read every chat, and trying to usually backfires. What you can do is lower the cost of telling you.

Announce a no-penalty rule, out loud, before anything happens

Say it plainly at a calm moment: “If you ever use ChatGPT or any app when you’re upset, you can tell me, and you won’t lose your phone for it. I’d just like to know how you’re doing.” Kids hide what they expect to be punished for. A rule announced in advance carries more weight than reassurance offered in the middle of a crisis.

Ask about the pattern, not the scariest message

A 2025 study by Yaman Yu and colleagues, a preprint that is not yet peer-reviewed, had parents and developmental psychology experts review real snippets of youth conversations with AI companions. Parents tended to flag single events as high risk. Experts looked for patterns, such as “repeated references to self harm or sustained dependence.” Ask how often, how long and what it helps with. Repetition over weeks tells you more than one alarming line.

Respond to disclosure with thanks first

If your teen tells you they have been talking to a chatbot about their feelings, your first words decide whether they tell you again. Start with “Thank you for telling me,” then “How has it been feeling lately?” Save your concerns about the app for later in the conversation.

Bring the doctor in on it

The JAMA Pediatrics authors say parents and clinicians should “proactively discuss chatbot use.” At your teen’s next checkup, mention it, or encourage your teen to. The APA recommends that users tell their health providers which AI tools they use.

What not to do: install secret monitoring as your first move

Covert surveillance can feel like safety, but if your teen finds it, you lose exactly the trust this article is about. If you are worried enough to monitor, you are worried enough to talk to a professional. Our guide to early warning signs and parent tools covers options that keep your teen in the loop.

What to Watch For Over the Next 3 Months

• Week 4: Have you said the no-penalty rule out loud? Notice whether your teen mentions AI use on their own afterward. A single unprompted mention is a good sign.
• Month 2 red flags: Hiding app use, long late-night chats, withdrawing from friends, changes in sleep or eating, or talk of being a burden or feeling trapped. The last two are among the warning signs of suicide listed by the National Institute of Mental Health. Any of these means calling a professional now.
• Month 3 self-check: Could your teen name an adult they would tell if a chatbot said something upsetting? Does their doctor know about their AI use?

Frequently Asked Questions

Is it bad if my teen asks ChatGPT for mental health advice?

Not automatically. Looking up what anxiety is, or rehearsing how to bring something up with a parent, can be reasonable. The concern is relying on it alone, in secret, for weeks. The APA says chatbots should not replace a qualified provider.

Why don’t kids tell their parents they use AI for this?

Teens in EdSource’s reporting cited fear of being judged, not wanting to worry anyone, and feeling there was no one to turn to. Some expect to lose their phone. Clear, advance reassurance that telling won’t be punished is the most direct counter.

My teen says the chatbot helped. Should I believe them?

Believe that it felt helpful. Then look at outcomes: sleep, school, friendships, mood over weeks. The RAND researchers note that high helpfulness ratings may partly reflect chatbots’ tendency to flatter users rather than the quality of the advice.

What should I do if I find chats about suicide or self-harm?

Talk to your teen today, calmly, and ask directly how they are and whether they are thinking about suicide; asking does not put the idea in their head. Call or text 988, text HOME to 741741, or use findahelpline.com. Call 911 if they are in immediate danger.

Are younger kids doing this too?

The two RAND surveys started at age 12, and use rose with age. The Common Sense census found 37% of AI users aged 9 to 17 had discussed feelings or personal problems with AI. Conversations about AI and emotions are worth starting before the teen years.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. McBain, R. K., Bozick, R., Diliberti, M., et al., & Yu, H. (2025). “Use of Generative AI for Mental Health Advice Among US Adolescents and Young Adults.” JAMA Network Open. https://doi.org/10.1001/jamanetworkopen.2025.42281
2. Brown University School of Public Health. (2025, November 18). “One in eight adolescents and young adults use AI chatbots for mental health advice.” Brown University. https://sph.brown.edu/news/2025-11-18/teens-ai-chatbots
3. McBain, R. K., Cantor, J. H., et al. (2026). “AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults.” JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2026.2015
4. RAND. (2026, June 1). “Nearly 1 in 5 U.S. Adolescents and Young Adults Use AI Chatbots for Mental Health Advice.” RAND Corporation. https://www.rand.org/news/press/2026/06/nearly-1-in-5-us-adolescents-and-young-adults-use-ai.html
5. Common Sense Media. (2026). “The Common Sense Media Census: AI Use by Tweens and Teens, 2026.” Common Sense Media. https://www.commonsensemedia.org/sites/default/files/research/report/2026-ai-use-by-tweens-and-teens-1.pdf
6. American Psychological Association. (2025, November). “APA Health Advisory on the Use of Generative AI Chatbots and Wellness Applications for Mental Health.” APA. https://www.apa.org/topics/artificial-intelligence-machine-learning/health-advisory-ai-chatbots-wellness-apps-mental-health.pdf
7. Common Sense Media. (2025, November 20). “Common Sense Media Finds Major AI Chatbots Unsafe for Teen Mental Health Support.” Common Sense Media. https://www.commonsensemedia.org/press-releases/common-sense-media-finds-major-ai-chatbots-unsafe-for-teen-mental-health-support
8. Seshadri, M. (2026, September 9). “‘I’m listening hard’: Teens increasingly turn to AI chatbots for mental health support.” EdSource. https://edsource.org/2026/teens-use-ai-mental-health/765573
9. Yu, Y., et al., & Wang, Y. (2025). “Principles of Safe AI Companions for Youth: Parent and Expert Perspectives.” arXiv (preprint, not peer-reviewed). https://arxiv.org/abs/2510.11185
10. National Institute of Mental Health. “Warning Signs of Suicide.” NIH. https://www.nimh.nih.gov/health/publications/warning-signs-of-suicide