How to Tell If a Photo or Video Has Been Faked by AI

Spot deepfakes and AI-generated images. What to look for, free tools to check, and when not to share.

Listen to this guide

Prefer audio? Tap play to hear an AI-narrated version.

Tap play to listen — audio is generated by AI narration.

How to Tell If a Photo or Video Has Been Faked by AI

A photo surfaces showing a local politician ‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍doing something shocking. A video clip goes viral of a celebrity saying something outrageous. A picture arrives in a family group chat that seems almost too perfectly timed.

In 2026, any of these could be real — or generated entirely by AI in seconds.

Knowing how to spot the difference is one of the most important media skills you can have right now.

Why This Matters

AI-generated images and videos — often called deepfakes or synthetic media — are now realistic enough to fool casual observers. They're used to:

Spread false information about politicians or public figures
Create fake "evidence" in scams and fraud
Generate misleading news stories
Impersonate real people in embarrassing or compromising situations

You don't have to become an expert. But knowing what to look for can stop you from sharing or acting on something that isn't real.

How to Spot an AI-Generated Photo

Look at the hands

AI still struggles with hands. Look closely at fingers — extra fingers, fused fingers, hands with unusual proportions, or fingers that blend into objects are a strong sign of AI generation.

Look at the background

AI-generated images often have backgrounds that are slightly blurry, contain repeating patterns that don't make sense, or include text that is garbled or unreadable.

Look at the teeth and eyes

Teeth may look too perfect, slightly misaligned, or unnaturally uniform. Eyes may have unusual reflections, asymmetry, or an artificial glassy quality.

Look at accessories

Glasses, earrings, and jewelry often look distorted, asymmetrical, or blend oddly into the face or hair.

Look at the text

If there's any text in the image — on signs, shirts, books, or in the background — AI often produces garbled, nonsense, or backwards text.

Quick test: Zoom into the hands and background. Those two areas fail most often in AI-generated images.

How to Spot an AI-Generated Video

Videos are harder to fake perfectly, but tell-tale signs include:

Unnatural blinking — too infrequent, too regular, or absent entirely
Facial edges that blur or shift — especially around the hairline and ears when the person moves
Lip sync that's slightly off — words and mouth movements that don't match precisely
Flat or inconsistent lighting — the face is lit differently from the rest of the scene
Unnatural head or eye movement — the head may move in a slightly robotic or limited way
Audio quality mismatch — the voice sounds crisp but the background audio sounds different

Tools That Can Help You Check

Several free tools can analyze images or videos for signs of AI manipulation:

Google Reverse Image Search — right-click any image on a computer and select Search image with Google to find where else it has appeared online. If it shows up in a completely different context, it may be misused or fake.
TinEye (tineye.com) — upload an image to search for its origins across the web
Hive Moderation (hivemoderation.com/deepfake-detection) — a free tool specifically for detecting AI-generated content
AI or Not (aiornot.com) — upload a photo and it will tell you whether it appears to be AI-generated

Questions to Ask Before Sharing Anything

Before sharing a photo, video, or news story that seems shocking or surprising, ask:

Where did this come from? Is it from a reputable news organization or an anonymous social media account?
Can I find this reported anywhere else? Real news events are covered by multiple sources.
Does something look slightly off? Trust your instincts — if it feels almost real but not quite, look closer.
Is this designed to make me angry or afraid? Emotionally charged content gets shared more, which is why manipulators use it.

The simplest rule: If something seems shocking and you only saw it in one place, wait before sharing it. Real stories get confirmed quickly by other sources.

What to Do If You've Shared Something That Turned Out to Be Fake

It happens to everyone — even journalists and researchers. If you realize you've shared a fake image or video:

Delete the post or message if possible
Send a follow-up to the same people letting them know it wasn't accurate
Don't be too hard on yourself — these tools are designed to deceive