Seeing Through AI: How Young People Can Spot Deepfakes Before They Share Them

Seeing is no longer believing. A video can show something that never happened. A voice can sound like someone you love. A headline can look convincing and still be false.

AI tools can create or change faces, voices, images, and video. You do not need special software to protect yourself. You need better habits, and you can learn them. Pause, check, and verify before you pass something along.

Understand What You’re Looking At

A deepfake is media that has been made or changed with AI to imitate a person or event. AI-generated media can also create something entirely new, such as a face, picture, voice recording, or video that never existed before.

Some fakes are made to entertain. Others are designed to mislead, embarrass, frighten, or scam people. And sometimes the media is real, but the caption gives it a false date or story. The goal is the same: get you to believe or share something without checking.

Why do these posts work? You naturally respond to faces and voices. You may trust a familiar voice or assume a video proves what happened. AI takes advantage of those instincts. Emotional posts can also travel quickly because anger, fear, and excitement make people want to react right away.

That reaction is your cue to slow down.

Check More Than the Picture

Look for clues, but don’t treat any single detail as proof. AI is improving, and real videos can look strange because of lighting, camera quality, or editing. A polished video can be fake, and a blurry one can be real.

Use these checks as starting points:

• Look closely. Do hands, teeth, jewelry, and other small details seem consistent? Check the edges of a face, where blur or odd blending may appear. Notice whether the background changes in a strange way.
• Listen carefully. Does the voice sound flat, oddly paced, or unlike the person’s usual speech? Do the lips appear to match the words?
• Check the account. Is it authentic and long-established, or newly created with little history? Does the account have a clear connection to the person, place, or event it claims to show?

These are clues, not verdicts. A fake may look convincing, and a real recording may have imperfections. Don’t decide based on appearance alone.

Trace the Post Back to Its Source

Before you share, ask who first posted the content. Look for the original footage, a named reporter, or a source with a clear connection to the event. Search the main claim and see whether trustworthy news outlets or official sources report it too.

Check the date. Old footage is often shared again with a new caption to make it seem current. A quick search for a distinctive phrase, place, or event may reveal that a clip appeared years earlier, or that no reliable source can confirm the story.

For images, try a reverse image search such as Google Lens. It can help you find other appearances of an image online. Remember: finding an older copy can reveal context, but search results alone do not prove whether a claim is true.

For video, look for the full version instead of relying on a short, cropped clip. A short segment may leave out what happened before or after. For practical verification ideas, see AFP Fact Check’s guide to finding a video’s source.

Pause Before You Share

If a clip makes you furious or euphoric, take a breath. That is precisely the moment to slow down. Ask what the post wants you to feel, and what it wants you to do next.

Maybe it urges you to repost immediately. Maybe it claims a loved one is in danger. Maybe it promises an unbelievable reveal. Whatever it is, don’t let urgency make the decision for you.

Use this simple rule: pause before you share. Spreading a fake can hurt the person shown, frighten a family, damage someone’s reputation, or help a scam reach more people. You don’t have to solve every mystery. You can choose not to amplify something you can’t verify.

Media-literacy guidance, including the UK government’s guide to deepfakes and media literacy, encourages people to assess sources and context rather than trust a first impression. That’s a useful habit for every kind of post, not only AI-made media.

Protect Your Family From Voice Scams

A call or voice message may sound like a loved one asking for money or personal information. The voice may seem familiar, and the story may sound urgent. Don’t assume that familiar means genuine.

Make a family plan before anyone is under pressure:

1. Choose a code word or verification question. Keep it private and agree how to use it.
2. Hang up and call back. Use a number you already know, not one supplied in the message.
3. Never act on urgency alone. Don’t send money, share passwords, or reveal private information until you’ve checked through another trusted channel.

If the person cannot answer your check, contact another family member or trusted adult. Take your time. A real loved one will understand that you are being careful.

Support Someone Who Gets Targeted

If a friend or family member is fooled, targeted, or impersonated, respond with care. Don’t mock them or blame them. Falling for a fake is not stupidity. It is being human in a moment when a tool is built to imitate humans.

Help them save the message, account name, link, and time it arrived. Encourage them to report the content on the platform and tell a trusted adult. If money, private information, or safety is involved, ask for help right away. You can also explore Minds Matter Resource Group’s mission and support or contact us for information about our community work.

If you shared something false, correct it publicly and simply. For example: “I shared this earlier, but I couldn’t verify it. Please don’t pass it on.” Then remove the original if you can. A clear correction helps limit harm. It also shows others that changing your mind is a strength.

Run This Quick Check

Before you trust a suspicious post, run through this list:

• Source: Who posted it first? Is the account authentic and connected to the event?
• Original: Can you find the full footage or a named report?
• Claim: Does another trustworthy source confirm the main details?
• Date: Could the image or video be old or taken out of context?
• Details: Do the face, hands, background, sound, and lip movements stay consistent?
• Action: Is the post pressuring you to share, send money, or reveal information?

Then ask yourself three questions before sharing anything:

1. Do I know where this came from?
2. Can I confirm the main claim somewhere trustworthy?
3. Could sharing it harm someone if it’s false?

If you can’t answer with confidence, wait. You are allowed to leave a post unshared while you check.

Be the Person Who Checks

You don’t need to catch every fake to make a difference. Start with one pause, one search, one call-back. Share what you learn with your family and friends without shaming anyone.

Maybe you’re the person who spots an old clip. Maybe you’re the one who asks for the original source. Whatever your role, your choices matter. Be the person in your group chat who checks before sharing, and help the people around you stay hard to fool.