Photo OSINT for Journalists: A Practical Workflow

The first pictures of a breaking story rarely come from photographers. They come from people who happened to be there, and they reach the newsroom through social media, messaging apps and reader tips, usually stripped of their metadata and screenshotted at least once. Before any of them can run, someone has to answer three questions. Is it real? Is it from where they say? Is it from when they say?

What follows is the routine many verification desks use in some form. None of it is complicated. The hard part is doing it every time, on deadline.

Keep the original

Before anything else, save the best version you can get and note where it came from: the account, the platform, when you found it and the claim attached to it. Posts get deleted, and you'll want that record later.

If someone sent you the image directly, ask for the original file rather than a screenshot. Sent by email or as a document, it may still carry its metadata.

Check whether it's old

Run a reverse image search on more than one engine and look for earlier copies. Recycled images are the most common problem by a long way. Real photos of past earthquakes, floods and protests get reshared every time a new one happens.

Read the metadata, if there is any

GPS coordinates and a capture time that fit the story are a good start. If there's nothing, which is normal for anything that went through a social network, move on.

Place it

Without coordinates, work from the image. Signs, language, license plates, architecture and landscape will usually get you to a country and often to a city. Then confirm it on satellite imagery and street-level photos by matching several details at once, such as the shape of a building, the shops beside it and the hills behind.

Check the time

The scene should fit the claimed date and hour. Compare the weather with historical records, look at the light and the season, and check the shadows if there's sun. Free browser plugins built for newsroom verification gather several of these checks in one place, video included.

Talk to the person who took it

Whenever you can, contact the source. Ask what happened, where exactly they were, and whether they have other photos or videos from the same moment. Ask for permission to use the image and how they'd like to be credited. A real eyewitness can usually answer in detail. Someone who found the picture online usually can't.

Think before publishing the location

Verification is there to get the story right, and it shouldn't put anyone at risk. A precise location in a photo of protesters, refugees or victims can expose them. Ask whether the exact spot is necessary for the story, and blur or crop what isn't.

Saving time on the routine part

Done by hand, these checks take anywhere from ten minutes to an hour per image, and a big story can bring in dozens. PixTrace runs the metadata check, the image search and the location analysis in parallel and shows the evidence behind each result. That leaves you more time for what needs a journalist: the source, the context and the decision to publish.

Try PixTrace on your own photos

Upload a photo and see where it was taken. 3 free analyses every month.

More articles