How Does Deepfakes: Manipulating audio and video to spread misinformation Work?

Deepfakes use tricks to change videos and sounds so they look like real people talking or acting, even if they're not.

Imagine you have a photo of your favorite cartoon character, and you want them to say something funny. You could draw their mouth moving while saying the words. That’s kind of what deepfakes do, but with real videos and sounds.

Deepfakes work by using computer tricks that look like solving a puzzle. They take many pictures or videos of someone's face, like your favorite cartoon character, and teach the computer how their face moves when they speak. Then, the computer can make it seem like they're saying new words, even if they never said them before.

It’s not just faces! The same kind of trick is used for voices too. If you record someone talking a lot, the computer can learn how their voice changes with different words and sounds. Then, it can make it look and sound like that person is saying anything, even if they never said it.

So next time you see a video or hear a voice that feels just right, remember: it might be a clever trick helping out!