· 9 min read
My Teen Talks to an AI Chatbot When Upset. Should I Worry?
Board-Certified Adult Psychiatrist and Child and Adolescent Psychiatrist
A 2026 survey linked some teen AI use for companionship or emotional help with more distress — a link, not proof of cause. Here’s how to respond, and when to get help.
Many teens use AI chatbots when they feel stressed or lonely. That is worth a calm, curious conversation — not a panic, and not a shrug. A 2026 Jed Foundation survey of 5,560 middle and high school students found that students who used AI for companionship, or who sought help from AI when they were struggling emotionally, were more likely to report anxiety, depression, loneliness, and a history of suicidal thoughts and attempts. That is a link, not proof that the chatbot caused those struggles. You can respond with curiosity, real-world connection, and a clear plan for safety.
If your teen seems to be turning to a screen instead of a person, the problem underneath is often loneliness. My Loneliness-to-Connection Toolkit gives you a step-by-step plan to help them rebuild real friendships.
What does the new data actually say?
The primary source is Exploring Artificial Intelligence [AI]: Understanding Relationships between AI Use and Mental Health Among Middle and High School Students, a Jed Foundation (JED) research report dated October 2026. Education Week covered it on November 10, 2026. JED analyzed its student survey, collected from October 2025 to April 2026, from 5,560 middle and high school students in 21 U.S. schools newly participating in JED programs that year. The data were weighted for non-response and were not nationally representative.
AI use was common: 81% of students reported using any AI. Fifty-eight percent used it for schoolwork, 28% for relationship or problem-solving advice, and 12% for friendship or companionship. Fifty-seven percent used AI when they felt sad, stressed, lonely, or overwhelmed, but only 8% said they used it “to get help” in those moments. Many were looking for distraction, advice, a way to calm down, or a way to feel less lonely, or turned to it because no one else was available.
Students who used AI for any reason did not, as a group, differ significantly in mental health from students who used no AI. The associations showed up with how AI was used. Students who used it for friendship or companionship, or who sought help from AI when emotionally struggling, had greater symptoms of anxiety and depression, more loneliness, a greater lifetime history of suicidal thoughts and attempts, and lower resilience and flourishing. A similar pattern appeared among students who used AI when they felt sad, stressed, lonely, or overwhelmed, and was generally strongest among those who had no one to talk to, were not ready to talk to a person, or used AI to feel less alone.
Kayla Anderson, JED’s vice president of research and evaluation and the report’s author, told Education Week that this is “very early research,” that it seems to matter what young people are doing with AI, and that some of these students may already be more vulnerable.
A separate study in JAMA Pediatrics, published online August 31, 2026, looked at 39,761 students in grades 4 through 12 in Ontario, Canada. About 21.1% reported using generative AI for emotional support or advice. That use was associated with higher emotional-problem scores and a higher rate of clinically significant symptoms, even after accounting for loneliness and a sense of mattering. The authors described this as a marker of distress, not proof of cause.
Does this mean the chatbot caused the problem?
No. Both studies are cross-sectional snapshots. They can show that two things occur together, but not which came first, or whether a third factor — such as existing loneliness — is driving both. JED states this limitation plainly: it is unclear whether certain uses of AI contribute to poorer well-being, or whether students who are already struggling are more likely to turn to AI in these ways.
I would not use these findings to scare a teen, and I would not use them to decide that AI is harmless. If your teen is leaning on a chatbot when upset, that is a signal. It may mean they need more human support or a closer look at mood, sleep, friendships, and safety. It does not, by itself, prove that the tool caused the distress, or that taking it away will solve it.
Why do teens turn to AI when they are upset?
Students in the JED survey often used AI in hard moments because it was available, because they wanted advice or a way to calm down, because they felt lonely, or because they were not ready to talk to a person. That is understandable. A chatbot can feel nonjudgmental when a young person is still trying to name what they feel.
The American Psychological Association’s June 3, 2025, health advisory on AI and adolescent well-being cautions that adolescents may be especially likely to trust a bot, to struggle to tell simulated empathy from genuine understanding, and to form relationships with AI systems that could interfere with real-world connection. In a November 13, 2025, advisory on generative AI chatbots and wellness apps, APA advised that these tools should not be used as a substitute for care from a qualified mental health professional. A chatbot is not a therapist, a parent, or a crisis service.
How can I start this conversation without making it a fight?
Curiosity usually opens more doors than a sudden ban. These sample questions are conversation starters, not a clinical script or a proven treatment: “What do you like about talking to it?” “When do you usually open it?” “Does it ever replace talking to a friend, or to me, or getting to sleep?” “If the bot wasn’t there, who would you want to go to?”
Listen for what the tool is doing for them, and notice what it might be crowding out: friends, sleep, or the adults they used to seek out. Then help them name two or three trusted humans they could go to on a hard day. The goal is to make sure a real person is also on the list.
When does this need more than a family conversation?
Please take it seriously if your teen’s mood, sleep, school participation, or friendships are slipping and staying that way; if they are withdrawing from people they used to trust; if the chatbot seems to be their only comfort; or if they talk about feeling hopeless, worthless, or better off gone. A qualified clinician can help you sort out what is going on. This article cannot diagnose your teen or tell you to start, stop, or change any treatment.
If your teen talks about suicide, self-harm, or being a danger to themselves or others, get help right away. In the United States, you can call or text 988 to reach the 988 Suicide & Crisis Lifeline, or chat at 988lifeline.org. The Lifeline is available 24 hours a day. If your child is in immediate danger, call 911.
Should our family tech agreements include AI tools?
Yes. AI tools now belong in the same conversation as phones, social media, and bedtime. A chatbot is not a therapist or a crisis line; 988 and a trusted adult are the plan for a true emergency. Using AI to brainstorm a school assignment is different from using it as a stand-in friend at 1 a.m. Privacy matters, because a teen may share thoughts they would not want stored or reused. Family rules should apply to adults too, and be revisited as your teen matures.
If you already have a media agreement, add AI to it. The point is not to ban every tool by default. JED’s findings suggest a more useful question: what is this tool being used for, and what might it be replacing?
The real question isn't the chatbot. It's who your teen has to talk to.
In the research above, the strongest warning signs showed up in kids who had no one else to talk to, or who used AI to feel less alone. Taking away the app doesn't fill that gap. People do.
I wrote The Loneliness-to-Connection Toolkit for parents who can see their child is "fine, but not quite fine." Inside you'll find:
- a four-week guided workbook you can start tonight
- a friendship-skills practice guide to help your child build connections that last
It's the same practical approach I use with families as a child psychiatrist, built so you can use it at home, at your own pace. You don't need to wait for a crisis to start.
Get the Loneliness-to-Connection Toolkit →
If your teen is in crisis, call or text 988 first. This toolkit is a support for families, not a replacement for care.
Frequently asked questions
Is an AI chatbot a therapist?
No. A chatbot can generate supportive-sounding language, but it is not a licensed clinician and should not be treated as mental health care. APA has advised that generative AI chatbots and wellness apps are not a substitute for care from a qualified professional.
Should I ban it?
Not automatically. A sudden ban can shut down the conversation and hide the distress underneath. Start by understanding why your teen is using it, what it is replacing, and whether there are safety concerns. Some families will reasonably set tighter limits, especially around late-night use or companion-style apps. The decision should fit your child, not a headline.
What if my teen says the AI helps?
Take that seriously. It may be helping them feel less alone, which tells you something about a need. You can acknowledge that and still ask whether a person in their life can share that role. A tool that feels useful can still be a signal that more human support is needed.
This article is for educational purposes and is not a substitute for medical advice, diagnosis, or treatment. It does not provide an individual diagnosis or recommend a change in medication or other care. Please consult a qualified clinician about your child’s specific situation.
Sources
- The Jed Foundation, Exploring Artificial Intelligence (AI): Understanding Relationships between AI Use and Mental Health Among Middle and High School Students, October 2026, https://jedfoundation.org/understanding-relationships-between-ai-mental-health-students-report, PDF: https://assets.jedfoundation.org/pdfs/AI_and_Mental_Health_Report_2026.pdf.
- Lauraine Langreo, “Students Who Turn to AI for Emotional Support Report Worse Mental Health,” Education Week, November 10, 2026, https://www.edweek.org/technology/students-who-turn-to-ai-for-emotional-support-report-worse-mental-health/2026/11.
- Kamilla Bonnesen, Amanda Krygsman, Sarah Hobson, Daphne Korczak, and Tracy Vaillancourt, “Affective Generative Artificial Intelligence Use and Youth Mental Health,” JAMA Pediatrics, published online August 31, 2026, https://jamanetwork.com/journals/jamapediatrics/article-abstract/2853453. (Abstract and University of Ottawa summary were reviewed; the full article text was not opened in this writing session.)
- American Psychological Association, “Artificial Intelligence and Adolescent Well-being: An APA Health Advisory,” June 3, 2025, https://www.apa.org/topics/artificial-intelligence-machine-learning/health-advisory-ai-adolescent-well-being; press release: https://www.apa.org/news/press/releases/2025/06/protect-adolescent-ai-users.
- American Psychological Association, “Artificial intelligence, wellness apps alone cannot solve mental health crisis,” November 13, 2025, https://www.apa.org/news/press/releases/2025/11/ai-wellness-apps-mental-health.
- 988 Suicide & Crisis Lifeline, https://988lifeline.org; SAMHSA, “988 Suicide & Crisis Lifeline,” https://www.samhsa.gov/mental-health/988.