How to spot AI videos on Facebook

To spot an AI video on Facebook, check who is behind it before you judge the picture: the Page's history, whether it's an ad, and who shared it with you. Then look for Meta's "AI info" label and scan the video's link. On Facebook the Page often tells you more than the pixels do.

Facebook is a little different from other apps. Many videos reach you from Pages you never followed, through suggested posts, groups and ads, or because a relative shared them. That's why the context matters so much.

1. Check the Page, not just the post

Tap the Page's name and look around before you believe the video.

• What else does it post? A Page that shares a stream of "heart-warming" or "shocking" clips several times a day, with no original reporting and no consistent theme, may be farming engagement. AI clips are cheap fuel for that.
• The caption formula. Watch for posts that ask you to do something rather than tell you something: "Type yes if you agree", "Share if this made you cry", "Nobody will like this". Comment bait helps a Page spread, whether the video is real or not.
• Its name and history. Some Pages show a Page transparency section with the date the Page was created, any name changes, and where its managers are based. Where it appears has changed over time and not every Page shows it, but if you find one, a recent creation date or a name that used to be something unrelated is worth noticing.
• Do the people in the videos ever repeat? A real farmer, rescuer or family appears again in other clips. Pages built on AI clips often show a different "real person" every time, none of whom you can find anywhere else.

2. Look for "AI info"

Meta labels some content its systems detect as AI-made, or that creators disclose, with an "AI info" label. At the time of writing, Meta has said the label is shown more visibly on content detected as fully AI-generated, and placed in the post's menu (the three dots) for content only edited with AI tools.

It's a helpful hint when it's there. But many AI clips carry no label, partly because re-uploads strip the hidden markers the label can rely on.

3. If it's an ad, check the advertiser

AI video turns up in Facebook ads that use a famous face to sell something: an investment "opportunity", a miracle cure, a government scheme that doesn't exist. These do real damage because they ask for money.

Before you click or pay:

• Tap the advertiser's name. Is it the real company or person, or a Page with a slightly different name?
• Meta's Ad Library lets anyone search for an advertiser and see the ads it runs. A Page running many near-identical ads with different celebrities is a strong warning sign. What the library shows can vary by country and type of ad.
• Check the celebrity's official accounts. If they really backed a product, it would usually be there.

Below is a frame from a clip of two business people shaking hands outside a bank, made with Kling 1.6. Generic, believable footage like this has nothing in it you could verify, which is why the advertiser matters more than the picture. Our video model scored it 99% AI.

Video VerifierKling 1.6

Video is likely

AI-made

Test set · Oct 2026crackly-ai.comVideo model: 99% AI

Our guide to deepfake investment ads and celebrity scams covers the money side in detail.

4. Look at the clip itself

Once you've checked the context, look at the video with a few targeted questions:

• Faces in emotional scenes. If the video's whole job is to make you cry or cheer, such as a soldier coming home or a child doing something remarkable, slow down. Watch the background people and the hands, not just the main face.
• Animals behaving perfectly. An animal that does exactly the touching thing the caption describes, with a smooth camera in the perfect spot, deserves a second look.
• Text in the scene. Signs, banners and uniforms can contain garbled letters in AI clips.
• The edges of the frame. Re-uploads are sometimes cropped, which can cut off a generator's watermark. A strangely tight frame is worth noticing.

5. When a relative shares it

Many people first see these clips because someone they love shared them. They meant well. If you find the video is AI, a private, friendly message usually goes down better than a public correction. Share what you found, such as a fact-check or a scan report, rather than just saying "this is fake". Our guide on telling your family a video is fake has wording that helps. The same approach works for videos that jump from Facebook to WhatsApp: see how to verify a WhatsApp forward.

6. Scan the video's link

Tap Share, then Copy link, and paste it into our Facebook video checker. A Quick Scan is free to try, needs no sign-up and answers in under a minute for most videos. For links, it also reads the post's caption and looks for published fact-checks, which helps with recycled clips.

Video in a private group or a friends-only post? No checker can fetch it without a login. Save the video if you can, or screen-record it, and upload the file instead.

What to do next

• AI and shown as real: don't share it. You can report the post to Facebook.
• An ad using a celebrity to sell investments: report the ad, and never pay or share personal details. If you've lost money, in India call the cybercrime helpline 1930 or report at cybercrime.gov.in.
• Unsure: treat it as unverified. If you share it, say so.

Quick answers

Why do I see so many AI videos on Facebook?

Suggested posts show you content from Pages you don't follow, and AI clips are cheap to make and get strong reactions. Engagement-focused Pages use them because they spread.

Does Facebook label AI videos?

Sometimes, with an "AI info" label. Many AI clips aren't labelled, so a missing label tells you nothing.