Fake images and videos are now easy to make and genuinely hard to spot. Teaching your child to recognise them may be the single most useful AI skill they learn this year. Here is how deepfakes work, and how to build a "spot the fake" reflex in a 10 to 16 year old.
Deepfakes are AI-generated or AI-altered images, audio and video that look real. The best defence is not a detection app — it is a habit: pause, question the source, and look for the tells before believing or sharing. This is a teachable skill, and children who build it early are far harder to fool.
What a deepfake actually is
A deepfake is media — a photo, a voice clip, a video — that AI has generated or altered to show something that never happened. The technology has improved fast, which is exactly why the old advice ("you'll be able to tell") no longer holds. Your child will see convincing fakes; the goal is to make them the sceptic in the room.
Why children are especially targeted
Young people share and reshare quickly, often trusting whatever appears in a feed. A fake image of a celebrity, a classmate, a news event or a "too good to be true" giveaway spreads because nobody paused to ask where it came from. Teaching the pause is teaching the defence.
Before believing or sharing anything striking, three questions: Where did this come from? Who benefits if I believe it? How could I check it? A child who asks these automatically is protected against most fakes — no app required.
How to spot a fake: practical tells
No single test is foolproof, but these help — especially in combination:
- Check the source first. A real event appears in more than one trusted place. A fake often traces back to one anonymous account.
- Look at hands, teeth, ears and backgrounds. AI still stumbles on fine detail — odd fingers, warped text, melted edges, mismatched shadows.
- Watch for too-perfect or too-shocking. Content engineered to make you feel something strongly is worth extra suspicion.
- Reverse-search the image. Show your child how to check whether a picture has appeared before, elsewhere, in a different context.
- Listen for audio glitches. Faked voices can have flat emotion, odd pacing or strange breaths.
The point is not to memorise a checklist forever — the tells change as the tech improves. The point is the underlying habit of checking before trusting, which outlasts any specific trick.
AI Fundamentals for Kids is a self-paced course for ages 10–16 — how AI works, how to prompt it, how to create with it, and how to stay safe and honest while doing it. No coding. You can preview every module yourself.
Explore AI Fundamentals for Kids → Ages 10–16 · No coding · Parent-guided · 7-day money-back guaranteeTurn it into a game
The most effective way to teach this is playful. Run a "real or fake?" session: pull up a mix of genuine and AI-generated images and have your child call each one, explaining why. They will get sharper fast, and they will start applying the same scrutiny to their own feeds without being told. The reflex to check before believing is, arguably, the most useful thing anyone can learn about AI right now.
The bottom line
Deepfakes are not going away, and detection tools will always lag behind. The durable protection is a child who instinctively questions the source, looks for the tells, and checks before sharing. Building that fake-spotting habit is a core part of AI Fundamentals for Kids — because it is one of the few skills that only grows more valuable over time.