When the chatbot becomes your child's best friend

Not because it is clever, but because it never mocks, never gets distracted, never interrupts. Those are things a teenager misses badly, and that is exactly where the worry starts.

AA Abdelilah Arahal
5 min read Updated 20 September 2026

I will be straight about the limits of what is known: this is a recent phenomenon and there is no long-run evidence yet. Anyone who tells you with certainty what it does over ten years is guessing.

But we can describe precisely why it happens, and that is enough to act on.

Why they are drawn to it

Not because it is clever. Because it:

Never mocks. A teenager fears judgement more than almost anything, and this listener never judges.

Is always there. Three in the morning, when nobody else is.

Is never distracted. It does not interrupt, does not glance at a phone, does not turn the subject back to itself.

Agrees a lot. And that is the part that genuinely concerns me.

What deserves attention

Not every conversation is a problem. A child asking about homework is not a teenager confiding. The signs worth a second look:

Substitution rather than addition. Contact with friends decreasing, rather than screen time increasing. The first is a signal; the second is ordinary.

Unusual secrecy. Closing the screen when you walk in, when they did not before.

Speaking of it as a person. Using a name, and "it told me" rather than "I read".

Distress when it is unavailable. A day without internet producing a disproportionate reaction.

Heavy subjects. Talking about sadness, fear, or harmful thoughts with a system rather than a person. That is not a sign about the screen. It is a sign that the human door is closed.

What to do

Do not ban abruptly. A sudden ban moves the use somewhere you cannot see and cuts the conversation at the moment you need it most.

Ask about content, not hours. "What do you two talk about?" is more useful than "how long?". The first opens; the second closes.

Show them it is a system. Sit together, ask it something, then ask the same question phrased the opposite way and let them watch it agree both times. Seeing that once does what ten warnings cannot.

Check the human door. If they are confiding in a system, the real question is not about the system: why are they not confiding in anyone? That is a question about you, their friends and their school, not about technology.

Get help when it is needed. If talk of self-harm or persistent sadness appears, that is a matter for a professional rather than a house rule. Do not hesitate and do not wait.

Voice specifically

Here is a view I will state plainly: voice tools are more dangerous for children than text ones.

Text has a gap in it. You read, you think, you reply. Voice removes the gap. It becomes a conversation, and conversation builds familiarity far faster than text on a screen ever does.

I have watched this happen: a child with a voice tool goes deeper. They talk more, they pause less, and they quickly forget they are speaking to a program.

So I will say something more definite than most articles do: if your child is young, keep them away from voice conversation specifically. Text is easier to supervise and slower to build a relationship.

What these companies are missing

And this is a demand I repeat every time I am asked:

We need real parental controls inside these tools themselves. Not a setting buried in a menu, but what we already expect from YouTube and from devices: a children's mode, limits on subjects, and a report to the parent about what is happening.

The second half of the demand matters just as much: teaching parents. Because a control nobody knows exists is a control that does not exist.

Until that arrives, the five rules above are what we have.

The other side

To be fair: some children, especially shy ones or those who find conversation difficult, find a safe rehearsal space here. Practising a hard conversation before having it is not a bad thing.

The difference between rehearsal and substitution is everything. Rehearsal ends in a real conversation; substitution cancels it.

In closing

We do not know the effect over ten years, and we know exactly why it happens now: because it does not judge, does not get distracted, and does not disagree. The remedy is not blocking the system. It is opening the human door.

Try this today: ask "what do you two talk about?" with genuine curiosity and no worried face. Their answer will tell you what you need to know.

Common questions

Why are teenagers drawn to talking with these systems?
Because they never mock, never get distracted, never interrupt, and are always available, which are things teenagers miss. The riskiest part is that they agree a lot, and a friend who always agrees is a mirror rather than a friend.
Which signs are worth worrying about?
Substitution rather than addition, meaning less contact with friends; unusual secrecy; speaking of it as a person; disproportionate distress when it is unavailable; and discussing heavy subjects with a system instead of a person.
Should I ban it?
Not abruptly. A sudden ban moves the use out of sight and cuts the conversation exactly when it is most needed. Ask about content rather than hours.
When should I seek professional help?
At any mention of self-harm or persistent sadness. That is a matter for a professional rather than a house rule, and it should not wait.
Share

No comments yet

Leave a comment

Never published. Used only if I reply to you directly.

Comments are read before they appear.

Related reading

12 min read

The best framework for building AI agents in 2026

Seventeen agent frameworks, ranked by what survives production rather than by GitHub stars. Some of the most popular names sit in the bottom tier, and one of them is a security incident waiting to happen.

Start with a short call

Fifteen minutes to understand what your team does and what you want to change. If training is not the right answer, I will say so.