AI-generated images are everywhere, and they’re getting better and better, making it harder and harder to know what’s real and what’s fake. Here are four tips for detecting images generated by AI.
Tip #1: Check the source
See who posted the content. Was it a reliable source, or an unknown user?
Look at the account. Does it regularly post images that are obviously created by AI?
Does the term “AI” appear in the account name or the bio?
Tip #2: Look at the details
Images generated by AI often contain bizarre errors in the background, such as hands with the wrong number of fingers, shadows that don’t make sense, or text that is deformed.
When in doubt, compare the image with real objects. For example, in a video that supposedly showed a Ukrainian soldier in tears because he was forced to join the army, the helmet contained errors. When you compare the helmet in the video with one used by the Ukrainian army, you can see differences, for instance an oval fitting that does not appear on the real model.
Tip #3: Look for AI watermarks
AI image generators like OpenAI’s Sora and Google’s Gemini insert an invisible watermark indicating what tool was used to create the image.
But users often blur or remove the watermarks. If you see a blurred zone where the watermark would normally be, it’s a sign the image was created by AI.
Tip #4: Use Google Lens
Put the suspect image into Google‘s reverse image search, Google Lens. It will tell you whether the image has been created or modified by one of Google’s AI tools.
This article was published to coincide with France’s Media in Schools Week, March 23-27, 2026.
