Skip to content
← All posts

· 8 min read

How to Talk to Your Child About AI-Generated Videos

By . Founder of ThinkaBit: a dad based in the Netherlands who has worked in tech for 25 years and now works with AI.

Part of our guide: How to teach your kids to spot deepfakes and AI fakes

Share:

A year or two ago, an AI-made video was easy to laugh at: melting hands, garbled signs, faces that slid around. That’s over. The clips in your child’s feed today can show an animal doing something impossible, a famous person saying something they never said, or a “news report” about an event that didn’t happen, and look entirely ordinary.

I notice it when I’m sitting next to my daughter. I’ll hear a voice and think, that’s generated. She just hears a voice. She isn’t being careless. Nobody has ever given her a reason to wonder.

This guide is about giving children that reason, without frightening them or turning every cartoon into a lesson.

Start with what they already watch

The worst way to begin is a serious sit-down about “the dangers of artificial intelligence”. The best way is on the sofa, in the middle of something they chose.

Wait for a clip that makes you curious and say it out loud:

  • “Hang on, do you think that really happened?”
  • “How do you think they filmed that?”
  • “That voice sounds a bit odd to me. What do you think?”

Then be quiet and let them answer. You’re not testing them, and you don’t need to know the answer yourself. “I’m not sure either, how could we find out?” is a perfectly good next line. You’re showing them what a thoughtful person does when something looks surprising.

Explain how it’s made, in one minute

Children are less easily fooled by things they understand. You don’t need the technical detail. Something like this is enough for a seven-year-old:

“Some videos are filmed with a camera. Others are made by a computer that has watched millions of videos and learned what things usually look like. You type what you want to see, and it draws a brand new video that looks real. Nobody filmed it, because it never happened.”

For a ten- or eleven-year-old, add the part that matters most:

“The computer doesn’t know what’s true. It only knows what looks right. So it can make a video of something that’s impossible, or of a real person doing something they never did.”

If they want to see it for themselves, make a harmless clip together with a tool on your own account, something silly like a cat driving a bus. Seeing how little effort it takes is more convincing than any explanation.

Three questions to teach instead of a checklist

We’ve written about the visual clues in how to spot a deepfake and how to spot AI-generated images, and they’re worth knowing. But the clues keep changing as the tools improve. Questions last longer.

1. Who posted this? A channel you know, a real news organisation, a friend? Or an account with a random name and thousands of similar clips? Where a video comes from tells you more than how it looks.

2. Why did they make it? To make you laugh, to sell something, to get you angry, to keep you watching? A video made to get a strong reaction deserves a second look.

3. Can I check it somewhere else? If something big really happened, more than one place will be talking about it. If the only evidence is one clip, be careful.

Turn these into a habit by asking them yourself, casually, a couple of times a week. Children copy what they hear far more than what they’re told.

Keep it calm

There are two ways this conversation goes wrong.

One is fear. A child who hears “you can’t believe anything online” either becomes anxious or, more often, decides the adult is exaggerating and stops listening.

The other is cynicism. If everything might be fake, then nothing matters, and real news gets dismissed along with the fakes.

The message to aim for sits between them: most of what you see is fine, some of it is made up, and you’re clever enough to tell which is which if you stop and think. That’s a skill to be proud of, not a threat to hide from.

It also helps to say that plenty of AI video is harmless fun. A dancing dinosaur doesn’t need fact-checking. The questions are for the moments that matter: when a video makes a claim about a real person, a real event, or something your child might act on.

When your child has been fooled

They will be. So will you. What you do in that moment decides whether they tell you next time.

  • Don’t laugh, and don’t say “I told you so.” Say “that one’s really convincing, I can see why you believed it.”
  • Share your own example. “I shared a photo last month that turned out to be fake.” Children are relieved to hear adults get caught too.
  • Work it out together. What made it believable? What would have given it away? Which of the three questions would have helped?

If they forwarded it to friends, help them put it right with a short message: “turns out that video wasn’t real.” Correcting yourself without embarrassment is one of the most useful habits you can show a child.

When the video is about someone they know

The harder case is a fake made to embarrass a classmate, or your child. Schools are meeting this more and more.

Make sure your child knows three things before it ever happens:

  1. Don’t share it, even to say how awful it is. Every share spreads it.
  2. Tell an adult straight away. A parent or a teacher. They will not be in trouble for what someone else made.
  3. Making a fake video of a real person to hurt them is serious. It isn’t a joke, and at their school it will be treated as bullying.

Say this once, plainly, and then leave it. The point is that they know where to go, not that they worry about it. We cover more of this in teaching kids online safety in the age of AI scams.

Make it a habit, not a talk

One conversation fades in a week. Small, regular moments stick.

  • Play “real or made up?” at dinner or in the car. Take turns describing a clip you saw and let the others guess, then explain why.
  • Let them catch you. Show them a video and admit you can’t tell. Being the one who spots it is a great feeling for a child.
  • Praise the question, not the answer. “Good thinking to check who posted that” matters more than whether they guessed right.

The aim is a child who, when you’re not sitting beside them, hears a small voice of their own asking who made this, and why?


ThinkaBit builds that habit for kids 7–12 with quick games like Fake or Fact, where they practise spotting what’s real and learn how AI makes the rest. It’s free to start: download it on the App Store or Google Play.

Share: