Using reverse image search to find copies of your photos

Reading the results

Results usually mix three things: pages showing your photo, pages showing edited versions of it, and pages showing different photos that look similar. Services tend to put visually similar images first, which is helpful for identifying an object and misleading when you are looking for copies.

Open each promising result and look at the image on the page itself, not the thumbnail in the results. Thumbnails are small and compressed, and two different photos of the same view can look identical at that size.

Telling a copy from a lookalike

Copyright protects your photograph, not its subject. Another photographer’s shot of the same lighthouse at the same hour is their own work. So the question is always whether the image on the page is your exposure, perhaps altered, or a different one.

• Look at incidental details. Clouds, waves, people in the background, reflections and shadows are effectively unique to one exposure.
• Check the framing. If one image is a crop of the other, everything inside the crop should line up exactly.
• Flip it. Mirrored copies are common. If nothing lines up, try comparing with your photo flipped horizontally.
• Compare files where you can. An identical file, byte for byte, settles the question. A matching distinctive file name is a strong hint but not proof, because files can be renamed.

What the results say about where to send a notice

The address of each result tells you what kind of place the copy is in. A social platform’s address means its own copyright form; a marketplace listing means its intellectual-property report; an ordinary website means its host’s designated agent, which our host lookup helps you find. The image file’s own address can point to a different server from the page, which is sometimes the one to notify.

How our photo search fits in

Our photo search runs the reverse image search for you and then checks the results against your file. Where it can fetch the image on the other page, it compares the bytes; where the file carries your photo’s distinctive name, it says so; and everything else is labeled a lookalike for you to compare. It leaves out your own sites and profiles if you list them, and it says plainly when the search provider did not answer, so an empty list is never mistaken for a clean one.

An uploaded photo is stored privately so the search can read it, then deleted with the search after 14 days, or sooner if you delete it from the results.

Why results change between searches

Running the same photo through the same service a week apart can give different results. Indexes are rebuilt constantly: new pages are crawled, removed pages drop out, and ranking changes which matches appear first. A copy you saw last month may have gone, or may simply have slipped out of the first page of results.

That is one more reason to record each copy when you find it, with a screenshot and the date, rather than relying on finding it again. It also means a search that finds nothing today is worth repeating later. Searching from a phone gives the same index as a desktop for most services, so use whichever lets you search your original file rather than a screenshot of it.

Location and language settings can shift results too, particularly for marketplaces and local businesses. If you sell or publish in more than one country, try the search with the settings for each.