How to tell if a photo has been edited

Photos are easy to change. A phone can remove a person, change a number or move a shadow in seconds, and AI tools can add things that were never there. If a photo matters, for a claim, a dispute, a sale or a news story, it's worth checking before you trust it. No single check catches everything, so use several together.

1. Look closely at the details

Open the photo at full size and go over it slowly. Edits often leave small mistakes that are easy to miss at a glance:

• Edges: blurry, too-sharp or haloed outlines around a person or object, often where it was cut out or pasted in.
• Light and shadows: shadows pointing different ways, or an object lit from a different side than everything around it.
• Repeats: the same patch of grass, brick or cloud appearing twice, a sign of a clone or healing tool.
• Bent lines: door frames, tiles or horizons that curve near a person, a sign of body-shaping filters.
• Text and numbers: letters that don't quite match the font, size or angle of the text around them.
• Reflections: mirrors, windows and glasses that don't show what they should.

Careful looking finds clumsy edits. Good edits can pass it, which is why the next checks matter.

2. Check the photo's hidden details

Most cameras save hidden details inside each photo, called metadata or EXIF data: the date and time, the phone or camera, its settings and often the location. Editing apps sometimes add their own name in the Software field. Our guide to seeing when and where a photo was taken shows how to read them on any device.

Treat metadata as a clue, not proof. It can be changed with free apps (the iPhone Photos app and Google Photos both let you change a photo's date), and WhatsApp, Facebook and most chat apps remove it completely when a photo is sent. Missing metadata doesn't mean a photo was edited, and metadata that looks right doesn't mean it wasn't.

3. Search for the photo online

A reverse image search finds other copies of a photo on the web. If an older or different version turns up, compare the two side by side. Try Google Lens, TinEye or Bing Visual Search. TinEye can sort results by date, which helps you find the first version.

This works best for photos that have been online before: viral pictures, product photos, news images. A private photo taken yesterday usually won't be found anywhere.

4. Ask for the original file

The simplest check is often the most effective: ask for the original, straight from the phone or camera that took it, sent as a file rather than through a chat app. An original has its full size and full details. If someone can only ever send a screenshot or a small compressed copy, be more careful.

5. Use a forensic tool, with care

Error level analysis (ELA) tools re-save a JPEG and show which areas compress differently from the rest, and pasted or retouched areas sometimes stand out. The results take practice to read: edges, text and bright areas can light up naturally, and a photo that has been re-saved a few times, as most shared photos are, shows very little. AI image detectors exist too, but they make mistakes in both directions, so never treat a detector's score as proof on its own.

6. Check a photo taken with a proof camera

All the checks above are detective work after the fact. The reliable way is to record the truth at the moment the photo is taken. Photos taken with the Timesnap camera carry the real time and place and a short code. At the same moment, the app stores a fingerprint of the photo: a grid of about 50,000 tiny squares, each with its brightness and colour.

To check one, drop the photo, or a copy someone sent you on WhatsApp, on timesnap.info. The copy is compared with its fingerprint square by square. Resizing and compression are allowed for, but real changes (an object removed, a number changed, something added with AI) are outlined in red, so you see exactly which part changed, not just a yes or no.

Have a Timesnap photo? Check it now, free.

Which check should you use?