Guide

How to tell if a photo is edited

Last updated

The only reliable way to know a photo wasn’t edited is a cryptographic signature added when it was taken, such as Content Credentials (C2PA) or a capture app’s seal. Without one, combine clues: get the original file, read its metadata, run a reverse image search, inspect edges and lighting, and check the context.

Quick checklist

  1. Ask for the original file. Request the file straight from the camera, not a screenshot or a forwarded copy. Messaging apps and social networks strip metadata and recompress images, which destroys most evidence either way.
  2. Check for a signed record. Look for Content Credentials (C2PA) or a capture app’s verification, such as a CaptureSeal QR code. A valid signature is the strongest evidence a file hasn’t changed since capture.
  3. Read the metadata. Look at the capture date, camera model and editing software in the file’s EXIF data. Missing or inconsistent metadata is a reason to look closer, not proof of editing.
  4. Run a reverse image search. Search the image with Google Lens, Bing Visual Search or TinEye to find earlier or uncropped versions.
  5. Inspect the image itself. Zoom in on edges, shadows, reflections, text and repeating textures. Look for smudged boundaries, mismatched lighting, warped straight lines and garbled lettering.
  6. Weigh the context. Compare the weather, daylight, signage and other photos from the same time and place. Ask who took it, when, and why it’s being shared.

1. Why the original file matters

Almost every check works better on the original. A screenshot, a WhatsApp forward or a social-media download has new pixels and usually no metadata, so it can neither prove nor disprove editing. If the photo matters (a deposit dispute, a claim, a news story), ask the sender for the file exactly as their camera saved it, by email, AirDrop or a file link.

2. Is there a signature or Content Credentials?

A growing number of cameras and apps sign photos when they’re taken:

A valid signature shows the file hasn’t changed since the signing device or app saved it. It doesn’t show the scene was real.

3. Read the metadata (EXIF)

On iPhone, open the photo in Photos and swipe up, or tap the info button. On a Mac, use Preview → Tools → Show Inspector. On Windows, right-click → Properties → Details. Look for:

Metadata is easy to strip and to fake, so its absence proves nothing and its presence proves little.

4. Reverse image search

Search with Google Lens, Bing Visual Search or TinEye. Finding the same scene earlier, at a higher resolution, uncropped or with different details is strong evidence of reuse or editing. Not finding anything doesn’t mean it’s original.

5. Inspect the image

Tools such as error level analysis can highlight areas compressed differently, but they’re easy to misread, and every re-save changes the result.

6. Check the context

Does the weather match records for that day and place? Do shop signs, cars and seasons fit the date? Is there another photo or video from the same moment? Who first shared it, and what do they gain? Many misleading photos are unedited but miscaptioned.

Proving your own photos weren’t edited: CaptureSeal

If you’re the one who needs to be believed, the most useful step happens before anyone asks: take the photo with a camera that signs it. CaptureSeal, a timestamp camera app coming soon to iPhone and Android, stamps the date, time, address and GPS coordinates into each photo and video and seals it with a key held in the phone’s secure hardware (Secure Enclave on iPhone, StrongBox/TEE on Android). Anyone can scan the QR code in the stamp, or drop the file on the verifier, to check in their browser that it matches the sealed original. An edited or re-saved copy won’t. It will be free for unlimited photos. See how it works, or get notified at launch.

What no method can prove

No check, signed or not, proves that what a photo shows is what someone claims. A real photo can show a staged scene or a screen. Use these tools to establish whether a file was altered, then judge the content on its context.

Questions

Frequently asked questions

Can you tell if a photo is edited from its metadata?

Sometimes. EXIF data can show editing software or a date that doesn’t fit, but metadata is easy to remove or rewrite, and most apps strip it when you share. Treat it as a clue, not proof.

Do AI image detectors work?

Not reliably enough to decide anything important on their own. They produce false positives on real photos and miss edited ones. A 2026 Veriff/Kantar survey of 3,000 adults found people’s own deepfake detection was “roughly a coin flip” (a vendor-commissioned study).

What is the most reliable way to prove a photo wasn’t edited?

Sign it at the moment of capture. Content Credentials (C2PA), Apple Reference Image on iPhone 18 Pro, and capture apps like CaptureSeal attach a cryptographic signature that breaks if the file changes. Inspection can only find signs of editing; a signature can confirm there wasn’t any.

Why does a forwarded WhatsApp photo fail verification?

WhatsApp and most messaging apps recompress images, which changes every byte even if the picture looks the same. Signature-based checks need the original file.

Can a real, unedited photo still be misleading?

Yes. A photo can be genuine but staged, cropped at capture, or shown with a false caption. No tool proves a scene is what someone says it is; context matters.

Coming soon to iPhone & Android

Take photos you can prove weren’t edited

CaptureSeal is coming soon to iPhone and Android. Leave your email and we’ll tell you the day it launches.

I’ll use it on (optional)