Offline Childhood
Search
← All conversations

Risks and guardrails10 of 13

Is that real?

A photograph is no longer evidence that something happened. Figuring out what is real is a skill now; it's no longer about something you can see.

How to explain it, by age

Same idea every time. Pick the version that fits for the child's age, and level up when it requires more complexity.

Ages 4-6

Introduce made-up pictures, which she already understands from drawing. "Computers can make pictures of things that never happened. They can look like photographs. If a picture looks strange or scary, come and ask me."

Try this

Generate a picture of something impossible but friendly, like your cat driving a bus. Making a fake together removes the mystery and the fear at the same time.

Listen for

Her asking whether a picture is real. The goal is for her to ask the question.

If the child already has more access than most: Add that a video can be made up too, which is less obvious to a child than a picture.

Ages 7-9

Extend to video and voice. "It can make videos of people saying things they never said, and copy someone's voice from a few seconds of recording. So seeing something is no longer proof it happened. The way you check is asking where it came from."

Try this

Take one photograph she has seen shared and trace it back to who first posted it. Actually doing it builds the habit better than being told to check.

Listen for

Her asking who posted it and where they got it, leaving the image itself for later.

If the child already has more access than most: Introduce reverse image search as a practical tool she can use herself.

Ages 10-12

Move the question from looking to sourcing. "You will not be able to keep up with how sophisticated this can get. The fakes improve every few months. What does keep working is asking where it came from, whether anyone reliable is also reporting it, and who benefits from you believing it."

Try this

Take something that made her angry or delighted online this week and trace it to an original source together. Notice if she has a strong feeling about it, and get her in the habit of connecting those feelings with the reality behind the content.

Listen for

Her noticing her own reaction and treating it as a prompt to verify.

If the child already has more access than most: Introduce the reverse problem: real footage dismissed as fake, which is increasingly the more common tactic.

Ages 13+

Name the deeper consequence, sometimes called the liar's dividend. Once everyone knows anything can be faked, genuine evidence can be waved away as fabricated. Beyond the false things that get believed, true things stop being provable, and that erosion is harder to repair than any individual fake.

Try this

Have him look at how newsrooms and organizations such as WITNESS verify material: provenance, metadata, corroboration, geolocation. It is investigative work, with perception barely involved, and seeing the actual method is clarifying.

Listen for

Him articulating why "you cannot trust anything" is the wrong conclusion and a dangerous one.

If the child already has more access than most: Discuss content provenance standards and why cryptographic signing of camera output is being pursued as a partial answer.

The habit outlasts the tells

Every article about spotting fakes lists visual giveaways. Those lists are obsolete within months, and worse, they leave a child confident in a skill that stopped working once the tech improved.

The durable skill is understanding the content’s origin. Ask: where did this come from? Who first published it? Is anyone reliable saying the same thing? Who gains if I believe it?

Those questions worked before synthetic media existed and will keep working after, because they are about the chain of custody rather than the pixels.

Do not land on total distrust

You don’t want a teenager who concludes that nothing can be known and therefore nothing is worth checking.

The goal is a child who can understand that fakery is cheap and common, that some sources are much more reliable than others, and that telling them apart is work worth doing.

Misconceptions to address

  • That you can spot fakes by looking closely. Counting fingers worked briefly and does not any more. Any list of visual tells has a short shelf life.

  • That only video and photographs are affected. Voice is the easiest to fake and the most convincing, because we have no habit of doubting it.

  • That this is only about famous people. The cheapest harm is ordinary: a fake voice note, a fabricated screenshot of a group chat, an image of a classmate.

  • Adults are considerably worse at this than teenagers, and considerably more confident.

Running this with a class

Show four images, two real and two generated, and have the class vote before revealing. Run it again a month later with fresh examples and compare the class accuracy. Watching their own detection rate fail to improve is the lesson: perception fails, so the answer has to be its origin.

The decision this connects to