Spotting AI fakes: voices, pictures, and video
A short clip of someone’s voice or a handful of photos is now enough to make a convincing fake. Spotting flaws is getting harder. Checking where something came from still works.
Read the transcript
The voice sounds like your child, and it wants money right now. Hang up. Call back on a number you know. Agree on a family code word. For pictures: who posted first? Who confirms it? Search the image. Notice what you feel. If someone fakes your child: don't pay, save evidence, report it. It's never the child's fault. Teach them to pause before sharing. safernotscared.org/ai/fakes.
The phone call that sounds like your child
Scammers use copied voices to fake a family emergency: a frightened voice, a demand for money, pressure to act now and tell no one. The Federal Trade Commission warns that a short audio clip, like one posted online, can be enough to clone a voice.
- Hang up and call your family member back on a number you already know.
- Agree a family code word that a real emergency call would include.
- Do not pay by gift card, wire transfer, or cryptocurrency because a caller says to.
- Tell grandparents about these calls. They are a common target.
Made-up pictures and videos
AI can put a real face on a fake body, make a photo of something that never happened, or animate a still picture. Clues like strange hands or garbled lettering are disappearing as the tools improve, so do not rely on them. Check where it came from instead.
Four checks before you believe or share
- Who posted it first? An account you know, a news organization, or a stranger?
- Does anyone else confirm it? Real events are usually reported in more than one place.
- Search the image itself. A reverse image search can show where a picture appeared before, and whether it was changed.
- Notice what it wants you to feel. Fakes are built to make you angry, scared, or rushed. A strong feeling is a reason to slow down.
When someone makes a fake of your child
Fake sexual images of real children, including ones made from ordinary photos, are abuse. The FBI has warned that ordinary photos and videos posted online are being turned into explicit fakes and used to threaten and extort. Sexual images of an identifiable child are illegal under federal law even when they are fake, and the child in the picture is never to blame. AI and the law has the details.
- Do not pay, and do not reply to threats.
- Save evidence: screenshots of the image, the account, and the messages.
- Report it to the NCMEC CyberTipline at report.cybertip.org or 1-800-843-5678.
- Use Take It Down to help remove nude or sexual images of anyone under 18 from participating platforms.
- Tell the school if classmates are involved.
The full first steps, in order, are in Someone is using photos to threaten us.
Teach the habit, not the fear
Kids do not need to spot every fake. They need one habit: pause before believing or sharing, and ask who made this and why. They also need to know they can come to you if someone makes a fake of them, without losing their phone.
Make your family a smaller target
- Keep accounts private where you can.
- Think before posting clear photos of your child’s face, or long videos of their voice, where anyone can see them.
- Agree a family code word, and make sure every adult who might get the call knows it.
Sources
- Scammers use AI to enhance their family emergency schemes — Federal Trade Commission
- Malicious actors manipulating photos and videos to create explicit content and sextortion schemes — FBI Internet Crime Complaint Center
- Take It Down — National Center for Missing & Exploited Children
- CyberTipline — National Center for Missing & Exploited Children