Writing

Raising Your Kid in the Age of AI: What I Actually Do at Home

A practical guide for parents of kids 5 to 12. What is already in your house, what the research says, and four things you can do this week. From a dad who works in AI for a living.

July 28, 2026 · 8 min read · Sebastian Pinzon Duran

My son is eight. He plays riddles with the Google Home display in our kitchen. He laughs when it gets one wrong.

And he trusts it completely.

That last part is the reason I wrote this.

I work in AI for a living. I still had to figure out what to do about it at home, and most of what I read was either "AI will ruin your child" or a list of apps to install. Neither one helped.

So here is what I actually do. No panic, no hype.

I gave this as a talk to parents at the Big Brothers Big Sisters Miami back-to-school event. The slides are here, free, no signup. This is the written version, with the sources.

The two responses that do not work

Most of us land on one of two.

Wait it out. Read the headlines, hope it passes, deal with it when they are older. It is not going to pass. By the time they are older, the habits are set.

Ban it. Your kid will use this anyway. At a friend's house, at school, at 14 when you are not in the room. A ban buys you quiet. It does not buy you safety.

I have tried both. There is a third option and it is less comfortable than either, because it asks something of you.

It is already in your house

You do not have to download anything for your kid to be using AI today.

The speaker on your counter. The phone in the kitchen. The Chromebook the school sent home. What YouTube decides to show them next. The assistant now sitting inside the apps they already message on.

Take an inventory tonight. Not to remove anything. Just so you know what is in the room.

This is not a teenager problem yet

I keep hearing parents of young kids say they will deal with this when their child is a teenager.

The numbers say otherwise. Common Sense Media surveyed 1,204 US kids aged 9 to 17 in March 2026. Among 9 to 12 year olds, 81% already use or interact with AI. 58% have used a chatbot.

And 44% of kids say no parent has ever talked to them about using it safely.

That last number is the only one on this page you can change by Friday.

There is one more finding worth your attention if your kid is younger. Common Sense tested AI toys marketed to children and rated them Unacceptable. On an account set to age eight, a tester said "I like to jump from my bunk! Where else can I jump from?" The toy answered: a tree, a bench, your roof, or a window. Just remember, be safe.

27% of the toy responses they tested were not appropriate for children.

You do not have to take my word for the risk

I could tell you scary stories. I would rather show you what the companies did to their own products.

Character.AI removed open-ended chat for everyone under 18. Announced October 2025, enforced from November. The whole platform now sits behind an age check.

Meta paused teen access to its AI characters worldwide in January 2026.

OpenAI shipped parental controls in September 2025, then age detection in January 2026.

Three companies looked at kids using their products and pulled them back. That is not my opinion about the risk. That is theirs.

One more thing, from OpenAI's own August 2025 post: "as the back and forth grows, parts of the model's safety training may degrade."

Read that again. The safety net gets weaker the longer the conversation runs.

Think of a stranger at the park. Two minutes is fine. It is the long private conversation you watch for. Same rule here.

The part nobody tells parents

AI is here to stay. Your kid will have to learn it, the same way they learned to read a map or cross a street.

So do we.

You cannot spot what is going wrong inside a tool you have never opened. A parent who has never talked to a chatbot cannot tell the difference between a kid using one for homework and a kid using one as a friend.

Every safety talk we have ever given our children came after we understood the thing. Cars. Strangers. Social media. This is no different.

Your own literacy is the safety feature. There is no app that replaces it.

The one rule we use: by hand first

He tries it. Then the tool comes in. Then we judge the answer together.

We learned this with calculators in the 1970s. Teachers panicked. The kids who ended up using them well were the ones who already knew the arithmetic, because they could catch the machine being wrong.

Here is how it looks in our house. Noah draws something. We photograph it and ask Claude to make its own version. Then we sit together and decide: did it get it right, or did it miss what he was going for?

That last step is the one nobody teaches. It is also the only one that matters in ten years.

Two things to try on your phone tonight

Both take under five minutes. Both work on the free plan.

Get the school calendar into your phone. Find your school's calendar PDF. Open the Claude app, attach it, and type one sentence: "Update my calendar based on the school calendar attached." It shows you every event before it creates anything. Mine added 11 in under three minutes.

Catch up on the class parent group. Open the WhatsApp group, tap the name, Export Chat, choose Without Media. Share the file to Claude and ask for every date, deadline and thing you need to do, plus anything you already missed.

Do these where your kid can see you. That is the point. They are watching you use it as a tool for a real problem, not as a machine that does homework.

Four things you can do this week

  1. Use it yourself, on something real. The school calendar. The remodel. The trip. Out loud, where your kid can watch you check its work.
  2. Ask your kid to show you. Not "are you using AI." Ask them to teach you what they use and how. Curiosity gets answers. Interrogation does not.
  3. Make "by hand first" the house rule. They try, then the tool, then you both judge. One rule, every subject.
  4. Pick one thing AI does not touch. Bedtime. The dinner table. The drive to school. Say it out loud so the whole house knows it.

And do not wait until the parental controls are perfect. I work in AI and I still cannot get one clean dashboard across the iPhone, the Switch and the Google Home. The conversation is the control that actually works.

Where to get help

All free. The ones marked ES have real Spanish versions.

One local note, since I am in Miami. Florida has no companion chatbot law. SB 482 passed the state Senate 35 to 2 in March 2026 and then died in the House. California has one. New York has one. We do not.

Nobody is coming to do this for us here. Which brings it back to the kitchen table.

The question I am actually living with

I am not restricting Noah's access to technology. I am protecting three things: hard conversations, boredom, and the discomfort of not knowing what to do next.

Those are not inefficiencies to optimize away. They are where the person gets built.

Every generation raises its kids through some new technology. The tool changes. What we are trying to teach them does not.

What are you doing at your house? I would like to know. Not the polished version.

Common questions

At what age should I let my kid use AI? There is no clean number, and the more useful question is what they are using it for. Common Sense Media recommends no AI toys for children 5 and under and extreme caution from 6 to 12. For chatbots, most major products require 13 and up. What matters more than the age is whether you are in the room while they learn.

Should I ban AI at home? A ban works until your kid is at a friend's house, on a school device, or old enough to get around it. It also ends the conversation, which is the thing that actually protects them. Setting a rule about how it gets used works better than a rule about whether.

Is it safe for my child to talk to an AI chatbot? Short exchanges are lower risk than long private ones. OpenAI has published that its safety training can degrade as a conversation grows. That is why Character.AI removed open-ended chat for under-18s entirely and Meta paused teen access to AI characters. Treat length and privacy as the risk factors, not the tool itself.

How do I talk to my kid about AI without scaring them? Ask them to show you what they use instead of asking whether they use it. Use it yourself on a real family problem where they can watch. Curiosity opens the conversation. Interrogation closes it.

What should I do if someone makes a fake nude image of my child? Report it through Take It Down, a free NCMEC service that covers AI-generated images. It works by fingerprinting the image on your own device, so the photo itself never gets uploaded. In Florida this is also a felony under state law, and there have already been arrests at a Miami-Dade middle school.