My Teen Talks to an AI Chatbot Instead of Me. Should I Be Worried?

Let's unwrap this...

Some worry is warranted. Panic is not. An AI companion your teen talks to is not automatically dangerous, and it is also not the harmless "it is just a tool" the marketing implies. What matters is which of two very different things is happening. Your kid might be using a chatbot now and then to think out loud, the way an earlier generation filled a journal. Or your kid might have made a machine the main place they take their inner life, and stopped bringing it to any human at all. The first is usually fine. Acting on the second, calmly and soon, is the point of this piece.

So the question is less about whether your teenager uses AI, and more about what the AI is replacing.

Teen talking to AI

Why your teen is telling a bot the things they will not tell you

Start by dropping the assumption that this is about you failing. A well-designed companion app is available at any hour, never tired, never disappointed, and never busy with your siblings or your job. It does not sigh, and it never brings up the last time you messed up. For a teenager whose nervous system reads most face-to-face conversations as high-stakes, that combination is powerful. They get to say the scary thing and feel none of the exposure that usually comes with it.

That appeal is real, and it is worth respecting rather than mocking. A kid who talks to a chatbot about a breakup or a friendship rupture is just reaching for the fastest available relief, which is exactly what the whole culture has trained all of us to do.

The concern from researchers is not hypothetical hand-wringing. Common Sense Media, which evaluates media and tech for families, reviewed the leading AI companion apps and rated them an unacceptable risk for anyone under eighteen, citing exposure to sexual content, harmful advice, and design that encourages emotional dependence. Their research also found that a majority of teenagers have already used one. In 2026, parents have filed wrongful-death and negligence lawsuits against companion-app makers, alleging that a chatbot deepened a child's crisis instead of interrupting it.

These tools can offer a lonely kid a moment of feeling heard. And the same tools are built to keep users engaged, which is a different goal from keeping users well. A product optimized for how long you stay is not the same as a friend who wants you to go outside.

What an AI companion cannot do, no matter how good it gets

I have spent years helping people unlearn the habit of outsourcing the hard parts of their lives, and I wrote a while ago about my unease watching kids hand their homework to AI.

Emotional support raises the stakes on that same worry.

A chatbot cannot feel the catch in your kid's voice that says the real problem is underneath the one they named. It cannot sit in a silence with them and let the silence do its work. Nothing in it will love them enough to say the thing they do not want to hear. And it never makes them tolerate the discomfort that grows a person.

Real relationships have friction. A friend you have to face again after a fight. The awkwardness of a parent who stays in the room while you are furious at them. Slowly finding out that another human can witness your worst and not leave. Those experiences are how a young person builds the nervous-system evidence that connection is survivable. A companion that removes all friction removes the training too. It hands your teen comfort while skipping the part where they learn they can handle a hard conversation and come out the other side.

How to tell ordinary use from a warning sign

You do not need to monitor every message to read the pattern. Ordinary use tends to sit alongside a life. Your kid mentions the app the way they mention a game or a group chat, still sees friends, still comes to you sometimes, still has bad days that pass. That is a teenager using a tool.

The signs worth your attention are about replacement and secrecy. Human relationships thinning out while screen time climbs. A kid who used to bring you things and has gone quiet, then you learn the confiding did not stop, it just moved to a place with no adult in it. Rising distress paired with more hours on the app rather than fewer. Talk of the bot as a best friend, a partner, or the only one who understands. Any mention of self-harm, in either direction of that conversation, is a stop-everything moment that calls for a professional, not a chatbot and not a lecture.

What to do as the parent when your teen talks to an AI chatbot instead of you

Do not open with a ban. A raid on the phone teaches your teen that the honest thing gets punished, which pushes the behavior further underground and confirms that you are one more person it is not safe to talk to.

Get curious first. Ask what they like about talking to it, what it gives them that people in their life do not. The answer is information you want. If it says the bot never judges them, you have just learned they are carrying a fear of judgment they have not felt safe bringing to a human. That is the actual thing to work on.

Then be the harder, slower, realer option on purpose. Stay in the room when it would be easier to fix it fast. Let conversations be awkward. Repair with them after a blowup instead of relitigating who was right. This is the ground my method, Exhale-Explore-Engage®, is built to walk families through: you interrupt the reflex to react or to control, you get underneath what is really driving the moment, and then you choose a response from that clarity. With AI in the picture, the work is the same one parents have always had, made more urgent by how good the easy alternative has become.

Set limits too, and set them as a family, out loud, the way you would agree on any household norm, rather than as a punishment aimed at one kid. Limits land better when they are shared agreements instead of verdicts handed down.

If your teenager is talking to an AI companion, you are not behind and you have not lost them. You have a signal. It is telling you something about what feels safe to them right now and what does not, and that is workable information. The goal is not to win them away from the machine by force. It is to make the human connection in their life feel worth the friction it costs.

If you want help doing that with your own teen, this is a large part of how I work with parents.

---

Masha Rusanov is a conflict and parenting coach, ICF-ACC credentialed, court mediator, and the author of Repatterned (Glass Spider Publishing).