How To Verify AI Images In Viral Posts

Viral images often look real and feel urgent. By the time you see one, it may have been shared millions of times, with the original source buried. Here's how to dig it out.

Trace it to the first post

1. Reverse image search with Google Lens, TinEye and Bing. TinEye can sort by oldest; Google shows "About this image" details.
2. Search the caption or key phrases in quotes.
3. Follow credits: "via" and "credit" mentions often lead back a few steps.
4. Check the earliest account: is it an AI art page, a satire account, or someone who was actually there?

Read the replies

Replies and quote posts on viral images often include people who've already traced the source or spotted problems. On X, Community Notes add context; on Meta's apps, look for "AI info" labels and fact-check overlays.

Look for fact-checks

Search the claim with "fact check." AFP Fact Check, Reuters Fact Check, AP, PolitiFact, Snopes and Full Fact regularly investigate viral images.

Spot recycled photos

Not every fake is AI. Old photos from different events are often reshared with new captions. Reverse search reveals them quickly: a "today" photo that appeared years ago is miscaptioned.

Visual checks

• Faces in crowds that blur or repeat.
• Text on signs and clothing that's garbled.
• Hands, limbs and objects that don't make sense.
• Light and shadows from different directions.

Metadata, if you can get the original

Viral copies have had their metadata stripped. If you find the original on a website or get the file from its creator, AI Image Detector (made by Patrick Bushe, who runs this site) can read its metadata in Chrome and flag AI tools or tags.

If you can't confirm

Don't share it as real. If you've already shared something false, correct it.

See a check before sharing and deepfakes in breaking news.