🎭 Deepfakes: Fakes Made with AI
6th Grade 🤖 Artificial Intelligence
A deepfake is a picture, video or recording that was made or changed with AI so that it looks or sounds like a real person did or said something they never did. The technology can be used for movies and jokes, but also for lies, scams and hurting people.
Where does the name come from?
The name joins two English words: “deep,” from “deep learning” — a kind of AI built on neural networks — and “fake.” The program learns from pictures, videos or recordings of a person what they look like, how they move and how they sound, and then it can produce new content that imitates them.
Faking pictures isn’t new — people were editing photos many years ago. What changed is that AI made faking faster, cheaper and sometimes very convincing, so even people with no special skills can do it.
Where do deepfakes show up?
In movies, similar technology is used to make actors look younger or to dub films into other languages. On social media there are funny videos where a person’s face is stuck onto a cartoon character. When everyone knows it’s just for fun, that can be okay.
The problem starts when the fake is meant to fool people: a video that makes it look like a politician said something they never said, a fake picture of a “disaster” that never happened, or an ad where a celebrity seems to recommend a product — without even knowing about it.
Voice scams: “Mom, send me money!”
One of the most dangerous uses is copying someone’s voice. Sometimes just a few short recordings of a person are enough to imitate how they sound. Scammers use this to call parents or grandparents in the voice of a child or grandchild and ask for money urgently.
How do you protect yourself? Agree on a secret family “code word” that only family members know. If someone calls in a panic asking for money or personal details, ask for the code word, or hang up and call that person yourself at the number you already know. Scammers try to create pressure — “Hurry, there’s no time!” — so stopping and checking is the best protection.
How do you spot a fake?
It used to be possible to spot a deepfake by signs like strange blinking, lips that don’t match the words, twisted hands or lighting that doesn’t make sense. Today fakes keep getting better, so it’s harder to rely on your eyes alone.
The important questions are: Who posted this? Are trustworthy news sources reporting the same thing? Is the video stirring up an especially strong feeling — anger, fear or laughter — so you’ll share it fast? If something looks too amazing or too shocking to be true, that’s the time to check before you believe it or pass it on.
There’s another problem worth knowing about: once people know fakes exist, they may start doubting real pictures and videos too, saying “that’s probably fake” about anything they don’t like. So the goal isn’t to suspect everything or believe everything, but to patiently check sources and look for confirmation from trustworthy ones.
Don’t make fakes of other people
Some apps let you “put” a person’s face onto a picture or video. Making a picture like that of a classmate, a teacher or anyone else without their permission — even “as a joke” — is hurtful and humiliating, and in some cases it’s also against the law. A picture like that can spread and hurt someone for a long time.
If you get or see a fake that hurts someone, don’t pass it on. Tell an adult — a parent, a teacher or the school counselor — who can help report it and get it taken down. And if someone made a fake of you, it’s not your fault, and it’s important to tell someone so you can get help.
Key facts
- A deepfake is fake content made with AI that imitates a real person.
- The name joins the English words “deep learning” and “fake.”
- Scammers use voice copying to ask for money urgently.
- A family code word helps you make sure you’re really talking to someone you know.
- Making a fake of another person without permission is hurtful, and in some cases it’s against the law.
Quick quiz
What is a deepfake?
Someone calls in a relative’s voice and urgently asks for money. What should you do?
Why do scammers say “Hurry, there’s no time!”?
What should you do when you get a fake that hurts a friend?
What does the “deep” in the name come from?
Is it okay to make a fake picture of a teacher “as a joke”?
Why is it harder today to spot a deepfake just by looking for clues in the picture?
Which use of similar technology can be okay?
💭 Think and talk about it
- Why do you think it’s easier to believe a video than written text?
- What code word would you pick for your family, and how would you explain to your grandparents why it matters?
- Who do you think should be responsible when a deepfake hurts someone — the person who made it, the people who spread it, or also the company that built the tool?
Questions kids ask
Is every edited photo a deepfake?
No. A color filter or cropping a photo is ordinary editing. A deepfake is a fake meant to make us believe a real person did or said something that didn’t happen.
Is there a program that can spot deepfakes for sure?
Some tools try to detect fakes, but none of them is perfect. So the safest way is to check the source and compare it with trustworthy sources.
What should I do if someone made a deepfake of me?
It’s not your fault. Save a screenshot as evidence, don’t respond to the person who did it, and tell a parent, teacher or school counselor right away so they can help report it and get it removed.