AI School › Staying safe online · Lesson 92 of 130

🎬 Manipulated videos

Videos can be faked too: the voice and the lips sometimes do not match.

A video seems like the safest proof in the world: if I see it, it happened. But today AI can change a person's face, make them say words they never said, or place them in a fake scene. These videos are called deepfakes. Not all of them are bad: some are jokes or movie effects. The trouble starts when someone uses them to trick people, to make others angry, or to ruin someone's reputation. That is why the golden rule works for video too: look carefully before you fully believe it.

There are signs that help. Watch the mouth: do the lips really move like the words you hear? Sometimes the movements arrive late or look glued on. Watch the eyes, which may blink too rarely, and the edges of the face, which sometimes flicker. Listen to the voice: if it sounds flat or strange, get suspicious. Finally, ask who posted the video and why. If you cannot find it anywhere trustworthy, wait. Do not share a video just because it is shocking: first verify, then decide calmly.

🎒 Try it in class (or at home)

Watch a short talking video and focus only on the lips: try to tell if they truly match the words.

← Lesson 91: Fake images and weird details · Lesson 93: Cloned voices →

📱 Download the Android app (beta) iOS coming soon

Say what you mean. Get what you need.

Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.