How to tell if a photo or video is AI-made
Can you tell for certain whether a photo or video was made by AI? Not always, and nobody can, not even experts. But you can get much better at spotting warning signs and checking where a picture came from. This guide shows you how, one small step at a time.
Step 1: Pause before you react or share
Fake images are often built to make you feel something fast: shock, anger, or delight. If a picture makes your heart jump, take a breath first. Slowing down is the most useful habit you can have.
Step 2: Look closely at the details
Zoom in and look for things that seem slightly off. Common clues include hands with odd fingers, teeth that look too even, jewelry or glasses that melt into skin, text on signs that is garbled, and backgrounds where things bend or repeat strangely. In videos, watch for blinking that looks unnatural, lips that do not quite match the words, or a face that flickers at the edges.
These clues are getting rarer as the tools improve, so a picture that looks perfect is not proof that it is real.
Step 3: Ask where it came from
Who posted it first? A news organization or a named photographer is more trustworthy than an account with no history. Try a reverse image search: in Google Chrome, right-click the picture and choose "Search image with Google." If the same photo shows up years ago with a different story, it has been reused.
Step 4: Ask an AI assistant to help you think it through
An AI assistant can't always look at a picture and give you a verdict, but it can walk you through a checklist and help you plan what to check. Try typing something like this:
"I saw a photo online of [describe it]. It was posted by an account I don't know. What are five things I can check to decide whether it is real, and where can I look them up?"
If your assistant lets you upload an image, you can add: "Here is the photo. Please point out anything that looks unusual, and remind me that you cannot be certain."
Step 5: Look for the same story elsewhere
If something big really happened, several reliable news sources will report it. Search for the event by name. If only one odd account is sharing it, be careful.
Step 6: Check for labels
Some apps and websites now mark images that were made or edited with AI. A label is a helpful sign, but a missing label does not prove a picture is real.
Step 7: Decide, and when unsure, do not share
You do not have to reach a final answer. "I'm not sure" is a perfectly good conclusion, and not passing it on is the kindest thing you can do for your friends and family.
What to watch out for
- AI detector tools and AI assistants can be wrong in both directions. They may call a real photo fake or miss a fake one. Treat their answer as one clue, never the final word.
- A picture that looks flawless is not automatically real, and a picture with flaws is not automatically fake. Older, blurry, or edited photos can look odd for ordinary reasons.
- Be extra careful with images tied to money requests, emergencies, or a "relative in trouble." If someone asks for money, hang up and call the person on a number you already know.
- For anything involving your health, legal matters, banking, or Medicare, do not rely on a picture or an AI answer. Confirm with your doctor, a lawyer, your bank, or the official website.
- Never click a link or enter personal details because of an image you cannot verify.
Want to try this with a friendly teacher?
Join a free live class on Zoom. Small group, no experience needed, no sign-up fee.
Reserve a free seat