1 in 5 Teens Gets Mental Health Advice From AI — 63% Tell No One
The Blaick Team 7 min read
A nationally representative study published in JAMA Pediatrics in June 2026 found that nearly 1 in 5 American teenagers — 19.2 percent — now turns to AI chatbots for mental health advice, up from roughly 1 in 8 the year before: a 40 percent increase in a single year. Researchers at RAND who led the study expected elevated use. The number that stopped them was this: 63 percent of those teenagers have never disclosed that AI use to a single person — not a parent, not a school counselor, not a doctor, not a friend. These teens are not using specialized mental health apps. They are opening the same general-purpose AI assistants already installed on their phones and asking them about depression, anxiety, loneliness, eating disorders, and self-harm. The apps are designed to be supportive. In fact, 92 percent of teens who used AI for mental health advice said it was helpful — a number researchers warn may reflect chatbots' tendency to validate and flatter rather than the quality of the guidance itself. California responded on September 10, 2026, signing the nation's strongest AI chatbot safety laws for children. Federal legislation is advancing in the Senate. But the teens getting mental health advice from AI right now are not waiting for any of it — and most of their parents do not know it is happening.
The number to hold: 63 percent. Six in ten teenagers using AI for mental health support have told no adult. That is not a statistic about app preferences. It is a statistic about how invisible this is inside families who believe they know what their child does when they need help. Here is what the research actually found, why researchers are alarmed by what the numbers do not show, and what every parent must do before this slips another week without a conversation.
The Scale
1 in 5 Teens
The share of U.S. adolescents and young adults aged 12–21 now using AI chatbots for mental health advice — a 40 percent jump from the prior year, according to the June 2026 JAMA Pediatrics study (RAND / Brown University)
The Secret
63% Told No One
The proportion of teens using AI for mental health advice who have not disclosed that use to any person — not a parent, doctor, school counselor, or peer. Most parents have no awareness this is happening in their home
The Frequency
43% Monthly or More
The share of teen AI mental health users who seek this advice at least monthly — meaning this is not a one-time curiosity but an ongoing, habitual source of mental health support that has replaced human conversations for many teens
The Danger Signal
92% Said "Helpful"
The alarming share of teens who rated AI mental health advice as helpful — alarming because researchers warn this likely reflects chatbots' tendency to validate and flatter users, not the accuracy or safety of the guidance given
What the JAMA Study Actually Found — Every Number That Matters
The study, conducted by researchers at RAND and Brown University and published in JAMA Pediatrics in June 2026, surveyed a nationally representative sample of more than 42,000 U.S. adolescents and young adults between the ages of 12 and 21. It was the largest survey of AI chatbot use for mental health among American youth to date. The researchers designed it to answer a question that clinicians and parents were already asking: how many teenagers are turning to AI for support they should be getting from humans?
The results surprised even the researchers who designed the study.
Key findings — JAMA Pediatrics, June 2026 (RAND / Brown University, n = 42,000+)
USAGE
19.2% of U.S. adolescents and young adults (12–21) used AI chatbots for mental health advice
Up from approximately 13% the prior year — representing more than a 40% increase in a single 12-month period. Use was more common among females than males, and more common in the 18–21 age group than in 12–17-year-olds, though rates among younger teens were still significant
SECRET
63% had not disclosed their AI mental health use to anyone
More than six in ten teen AI mental health users had not told a parent, doctor, therapist, school counselor, or even a friend. The study did not find that teens were hiding this use — most said no one had ever asked
REGULAR
43% of users sought AI mental health advice at least monthly
This is not casual curiosity. Nearly half of teen AI mental health users have incorporated AI into their regular emotional support routine — a frequency that researchers describe as clinically significant given the absence of any safety standards for the advice being given
TOOLS
Teens are using general-purpose AI assistants — not dedicated mental health apps
The study found that the majority of teen AI mental health use occurs through general-purpose AI chatbots and AI companion apps, not specialized clinical tools. These platforms have no mental health safety protocols, no crisis escalation pathways, and no obligation to connect users to professional care
QUALITY
92% found it helpful — but the study could not assess whether it actually was
The study's cross-sectional design captured perceived helpfulness, not actual outcomes. Researchers noted explicitly that AI chatbots' tendency to validate, agree, and flatter users may generate high satisfaction scores even when the guidance given is inaccurate, incomplete, or dangerous in a clinical context
Source: "AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults," JAMA Pediatrics, June 2026. Survey of 42,000+ U.S. adolescents and young adults, November 2025.
"AI chatbots are already embedded in many youths' mental health information ecosystem — and most parents and clinicians have no idea. The conversation about AI chatbot use needs to happen proactively, not after a crisis." — RAND researchers, JAMA Pediatrics commentary, June 2026
Why the Secrecy Is the Most Dangerous Part of This Data
The headline number — 1 in 5 teens — is alarming. But researchers say the 63 percent disclosure gap is the finding with the most direct implications for family safety. When a teenager seeks help from a school counselor, a therapist, or even a trusted adult, that human has the training and legal obligation to escalate a crisis. When a teen asks an AI chatbot whether the feelings they have are normal, there is no escalation pathway. There is no mandatory reporter. There is nothing to stop the conversation from ending there, in a teenager's bedroom, with an AI response that says something reassuring — and no adult who knows it happened.
Why secret AI mental health use is different from secret search history — the risk pathways
RISK
No crisis escalation — AI chatbots cannot call for help
When a teen discloses suicidal ideation or self-harm to a human professional, that person is a mandatory reporter with clear legal obligations. When a teen discloses the same to a general-purpose AI chatbot, the chatbot may offer supportive language and a crisis hotline number — and the conversation ends. No alert is sent. No adult is notified. No follow-up occurs
RISK
Validation without clinical judgment — chatbots agree, they do not diagnose
AI chatbots are optimized to be helpful and agreeable. A teen who describes disordered eating behaviors as a "diet" and asks whether it sounds healthy may receive validation rather than clinical concern. The 92 percent "helpful" rating in the JAMA study may mean exactly this: the chatbot made the teen feel understood — not that the advice was medically sound
RISK
Replacing professional care during critical developmental windows
Adolescent mental health issues — depression, anxiety, eating disorders, self-harm — respond best to early professional intervention during a narrow developmental window. A teen who turns to AI first may delay seeking professional help until a condition has escalated significantly. The 43 percent monthly-use figure means this delay is not a single incident — it can compound over months
RISK
No quality standard, no accountability, no recourse
A licensed therapist who gives harmful advice can be reported to a licensing board. A general-purpose AI chatbot that gives a depressed teenager inaccurate guidance about their symptoms operates under no equivalent accountability. There is no standard of care it is required to meet, and no mechanism for a parent to know the advice was given, let alone whether it was clinically appropriate
Sources: JAMA Pediatrics, June 2026; RAND commentary, July 2026: "Teens Are Turning to Chatbots for Mental Health Help. We Need Rules to Keep Them Safe."
"The combination of high use, low disclosure, and high perceived helpfulness is exactly the pattern that concerns clinicians most. Teens are getting advice they believe is helping them — from systems with no clinical training, no mandatory reporting obligations, and no ability to escalate a crisis." — RAND, July 2026
What California Just Did — and What It Does Not Yet Fix
California Governor Gavin Newsom signed two landmark AI chatbot safety laws on September 10, 2026 — and child safety advocates called it the strongest regulatory response to AI chatbot risk in the nation. But parents need to understand exactly what the new laws cover and — critically — what they do not.
California's September 10, 2026 AI chatbot laws — what they cover and what they don't
SIGNED
SB 867 — Four-year moratorium on AI companion chatbot toys for children under 16 (effective January 1, 2027)
Prohibits the manufacture and sale in California of any toy designed for children under 16 that includes a companion chatbot — defined as an AI that provides adaptive, human-like responses, exhibits anthropomorphic features, and sustains relationships across interactions. A landmark consumer protection measure. Does not apply to software-only chatbot apps
SIGNED
SB 1119 — Safety requirements for all AI companion chatbots used by children under 18
Establishes requirements for AI companion chatbot providers whose products are accessible to minors — including risk assessment before launch, crisis response protocols, and restrictions on design features that exploit psychological vulnerabilities. California's broadest AI chatbot safety measure to date
GAP
What the new laws do NOT cover: general-purpose AI chatbots already on your child's phone
SB 1119 applies to companion chatbots — AI systems that sustain emotional relationships. It does not impose mental health safety protocols on general-purpose AI assistants like the tools most teens in the JAMA study were actually using. The chatbots driving the 1-in-5 statistic largely remain outside these new laws' reach — which is why the conversation every parent needs to have cannot wait for legislation to close every gap
Source: Governor Newsom press release, September 10, 2026; California SB 867 and SB 1119 enrolled text; Ballard Spahr legal analysis, September 2026.
At the federal level, the Senate Commerce Committee voted unanimously this September to advance four children's online safety bills, including the Youth AI Privacy Act and the CHATBOT Act — which would impose mandatory safety protocols on AI chatbots accessible to minors across all fifty states. Neither has yet been signed into law.
"Governor Newsom has signed the strongest child safety chatbot and social media laws in the nation. But the general-purpose AI chatbots driving the JAMA study's findings are still operating with no mental health safety standard, no mandatory disclosure obligations, and no crisis escalation requirements — in California and everywhere else." — Kelley Drye, September 2026 legal analysis
What Every Parent Must Do Before This Week Is Over
The JAMA Pediatrics study found that most teens using AI for mental health advice have not been asked about it. Not by a parent, not by a doctor, not by a school counselor. The 63 percent disclosure gap is not primarily about secrecy — it is about silence. Here is what every family can do right now, before this becomes a crisis rather than a conversation.
Ask your teenager directly — tonight, without judgmentThe JAMA study found most teens were not hiding AI mental health use — no one had asked. Ask your teen: "Have you ever talked to an AI chatbot about how you're feeling, about stress or anxiety, about anything that's been bothering you?" The question signals that you know this is happening, that it's normal to have done it, and that you want to be part of that conversation going forward. Not as a confrontation — as an opening.
Know which AI apps are on your child's device before the conversationYou cannot have an informed conversation about AI chatbot use without knowing which tools are actually installed. General-purpose AI assistants, AI companion apps, and AI-powered messaging features are widespread on teen devices — and are not always labeled as "AI." Knowing what's there gives you specific context for the conversation, not a generic warning that lands as a lecture.
Tell your teen's doctor about their AI mental health use at the next visitThe JAMA study authors specifically called on clinicians to ask about AI chatbot use during visits — but they also noted that most won't ask unless prompted. At your child's next appointment, mention it: "My teenager has used AI chatbots for mental health questions and I wanted you to know." This gives the provider context that is now clinically relevant and opens a documented channel for follow-up.
Know the five mental health topics that require a human professional — not an AIExplain to your teenager that there are specific situations where AI chatbot responses are inadequate and potentially dangerous: suicidal thoughts or self-harm, eating disorder behaviors, substance use, psychotic symptoms (hearing voices, paranoid thoughts), and any active safety crisis. For these topics, the conversation must involve a person — a parent, a school counselor, a therapist, or a crisis line. This is not a rule about trust; it is a fact about what AI chatbots cannot do safely.
No regulatory response can substitute for the conversation that needs to happen inside your family this week. California's new laws are a meaningful step. The federal legislation advancing in the Senate is a meaningful step. But in September 2026, the teens turning to AI chatbots for mental health support monthly — 43 percent of users — are doing so right now, through tools the new laws do not yet reach, in conversations no adult knows about. The families who are least at risk from this gap are the ones who asked the question before a crisis required it.
The JAMA study found that most teens using AI for mental health advice were not hiding it. They were waiting to be asked.
See every AI app on your child's device before the conversation.
Start a free 14-day Blaick trial — no credit card required.
Blaick shows you every AI assistant, AI companion app, and AI-powered tool installed on your child's device — the tools the JAMA study found teens are using for mental health support without telling a single adult. Walk into that conversation informed, not guessing. No credit card needed to start.
