This is a contributed piece from Scott Kosch, a dad of two and the author of Indispensable: A Student’s Guide for Thriving in the Age of AI.
Picture this: your kid had a rough day. Something happened with a friend, a coach, or an exam grade. They came home, went to their room, and talked it through.
Just not with you. And not with a friend either. With a chatbot.
The Data That Filled a Room in Washington
At a Capitol Hill briefing in July, The Jed Foundation, a leading nonprofit in teen mental health, presented research from a survey of 5,500 middle and high school students on AI use.
Some of the findings you could guess. Four in five students use AI. More than half use it for schoolwork. Nearly a third use AI for relationship or personal advice. More than one in ten talk to AI for companionship.
The mic-drop statistic: nearly three in five students have used AI when they felt sad, stressed, or lonely.
There’s one more detail, buried deeper in the data, that tween parents should read twice: middle schoolers are more likely than high schoolers to reach for AI in those low moments. The kids still learning to swim are the ones drifting farthest from the lifeguard.
Why Your Kid Might Use AI for Companionship (and Why It Makes Sense)
Before you panic or confiscate anything, understand the appeal. A chatbot is available at 2 a.m. when everyone else is asleep. It never judges. It never gets tired of listening to the same problems. It never changes the subject, and (this one matters) it never tells you.
It’s basically a golden retriever that can spell.
Your kid isn’t broken for finding that appealing. If you sat down to design the perfect confidant for a self-conscious 12-year-old, you’d build ChatGPT.
AI chatbots are trained to keep users engaged. The easiest way to keep someone talking? Agree with them. Tell a chatbot someone wronged you, and it will usually take your side. A good friend sometimes says, “Actually, I think you owe her an apology.” A chatbot rarely does. It’s a mirror that only shows your kid what they want to see.
Psychologist Jonathan Haidt was asked what worries him most about teens and AI. You’d expect him to say screen time. Instead, he said: “The most dangerous thing seems to be the relationships, the long conversations.”
Kids today are the first generation this is happening to. They’re the test subjects, just like the last generation was for mobile phones and social media. That’s reason enough to act now.
You may also like to read: Technology Can Make Your Teen Anxious And Depressed, Here’s How To Help
Five Things Parents Can Do Today Regarding AI and Teens
Here’s what to actually do.
1. Swap the cheating question for the curiosity question.
Most parents’ only AI conversation with their kid is about homework. That guarantees you’ll never hear about the 2 a.m. conversations. Instead, ask this at dinner, in a curious tone: “What’s the most interesting thing you asked AI this week?”
Then share yours. You want AI to be a normal topic in your house, like movies or music, so that when the emotional stuff comes up, it isn’t a confession. My son’s answer last week involved an analysis Claude did of historical fantasy football trade values.
My response? “Wow, that’s interesting. I’d like to learn more about how you did that.”
Interrogation closes doors. Curiosity props them open.
2. Teach them the sorting question.
Skip the “is AI good or bad” debate. That’s like debating whether fire is good or bad. (Cooking dinner? Great. Cooking your house? Less great.) Instead, ask: “Is AI helping you connect with people or replacing them?”
And give examples:
- Using a chatbot to rehearse a hard conversation with a friend before having it? Tool.
- Using AI instead of ever having that conversation? Crutch.
- Venting to AI at 2 a.m. to fall asleep, then telling a real person the next day? Tool.
- The 2 a.m. conversation becoming the only conversation? Crutch.
As teen psychologist Lisa Damour explains: “Avoidance feeds anxiety.” Every hard conversation a kid dodges with a chatbot makes the next one harder.
3. Help them name their human — and accept it might not be you.
The best guardrail I’ve found is what I call Name Your Human. Your kid picks one person — a friend, a parent, a coach, a counselor — who also gets to hear the heavy stuff, even if a chatbot sometimes hears it first.
Here’s the part that stings: their human might not be you. That’s okay. A teen who talks with their aunt or coach is a teen talking to a real person. Your job is to make asking feel easier for them.
If you are their person, remember that telling you might feel riskier than talking to a chatbot. React calmly to small disclosures if you want to earn the big ones.
4. Watch your middle schooler more closely than your high schooler.
Most parents assume AI risk scales with age and independence. The data suggests the opposite: middle schoolers are the heaviest users of AI for emotional issues. So, flip your assumptions.
For kids roughly 10 to 14, treat AI chat like you (hopefully) treated early social media. Use it in common areas of the house rather than behind a closed door, on devices you can see, and have regular low-stakes conversations about it. A 12-year-old confiding in a chatbot needs more attention than a 17-year-old using one to study for the SAT.
You may also like to read: Here Are The Tech Rules You Need To Be Setting For Your Tweens and Teens
5. Draw one bright line, out loud.
Say this to your kid: “If you are ever thinking about hurting yourself, a chatbot is the wrong door. Call or text 988, or come get me at any hour, no matter what.”
More adults should be saying this to the kids in their lives. Beyond the practical instruction in this message, a kid hears, “I care about what happens to you.” No kid can hear that from too many people.
Even the student panelists in Washington made this point: at its best, AI should point kids toward resources like 988. It should never replace them.
Kids follow bright lines better than vague warnings, but only if you actually draw them. Even if they roll their eyes, trust me, they still get that you care.
The Good News about Teens and AI
The students who spoke on Capitol Hill asked Congress for guardrails. But Washington moves like the Potomac River it sits on: slow and half the time backwards.
But the most important guardrails were never going to come from Congress anyway. They come from your dinner table: one curious question, one sorting test, one named human, one age-appropriate boundary, and one bright line. You now have a plan.
AI may be there for your kid at 2 a.m., and that’s okay. Sometimes, it will give them what they need in that late-night moment. Your job is making sure a human is available the next day — and that your child knows when AI is the wrong door.
If you or your child is struggling, call or text 988 (Suicide & Crisis Lifeline). It’s free, confidential, and available 24/7.
For more tips and tricks to help you navigate these tumultuous years of raising teens in a tech-savvy world, check out Scott Kosch’s book Indispensable: A Student’s Guide for Thriving in the Age of AI.
Parenting teens and tweens is a tough job, but you’re not alone. These posts might help:
5 Must-Have Phrases Every Parent of Teens Needs to Shut Arguments Down
Hilariously Funny Podcasts for Teens That Will Have Them Laughing Out Loud
*This post may contain affiliate links where we earn a small commission for purchases made from our site.
