How to Spot AI-Generated Images: A Fact-Checker's Guide

Fact-checkers have long been aware of the challenges posed by fake images online, but with the rise of generative artificial intelligence tools in recent years, this task has become increasingly difficult. At Snopes, we’ve spent a lot of time figuring out whether an image shared online was legitimate or created by AI. Determining whether an image is real often comes down to looking at small details for things that are out of the ordinary.

As generative AI tools have become more advanced, this task has become increasingly difficult, and even experts struggle at times. Ensuring that images you see online are real is a worthwhile skill to practice, especially given the skyrocketing amount of AI-generated content we encounter online. Below, we’ve compiled a list of common strategies Snopes reporters use to figure out whether an image is AI-generated.

These strategies are not foolproof – an image having any one of these characteristics does not immediately mean it’s a fake. We recommend using the tips as a checklist instead. The more signs on this list you see in an image, the more likely it is to be AI-generated. As generative models continue to improve, some of these tips and tricks might become obsolete.

One common issue with AI-generated images is their depiction of light effects. Generative AI models often struggle to accurately depict real light and all its effects. The following tips revolve around light effects that Snopes’ team has noticed AI models often struggle to get correct.