How to Find Out Who Leaked Your Photos
Short answer: If you shared a photo and it showed up somewhere it should not, you usually cannot tell who leaked it from the file alone, because every recipient got an identical copy. The way to make leaks traceable is to give each person their own copy with a different invisible mark, before you send anything. When a copy leaks, you upload it (even a screenshot) and the mark tells you whose copy it was. This only works for copies that were marked beforehand.
Last updated: October 9, 2026.
Why you usually cannot tell who leaked a photo
If you send the same file to ten buyers, clients or friends, all ten copies are identical. Once one of them is reposted, nothing in the image says which person it came from. Metadata such as EXIF is stripped by most platforms and disappears with a screenshot. A visible watermark can be cropped or edited out, and it is the same for everyone unless you manually make ten versions.
What a per-recipient invisible watermark does
A forensic (invisible) watermark changes the pixels of an image in a way that is not meant to be noticed when you look at it. Each recipient gets a copy with a different identifier hidden in it. Later, a detector reads the identifier back from whatever file you find.
Markedfile is designed so the mark survives common things that happen to leaked images: screenshots, recompression, crops as small as roughly 230 px of a detailed area (smaller crops can fail) and most 90° rotations. Detection is heuristic, so there is no 100% guarantee. A heavily edited or very small crop may not be identifiable.
Mark
Share
Leak
Detect
How to do it with Markedfile
- Before sharing, upload the original photo and create one marked copy per recipient (for example: Ana, Bruno, Carla). Use any label you will recognise later.
- Send each person their own copy through the channel you already use: email, WhatsApp, Drive, Telegram, a download link.
- If a photo turns up where it should not, save it or take a screenshot of it.
- Upload it to Detect. Markedfile reports which recipient's copy it matches.
- Decide what to do. The result tells you which copy leaked. What you do with it (a conversation, cutting off access, a takedown request) is up to you.
Files are processed in memory and are not stored. The Free plan covers 5 active files with detection for 6 months, so you can try it on a real set before paying.
What this cannot do
- It cannot trace copies you sent before you used it. It only recognises copies Markedfile marked.
- It does not prevent a leak. It tells you which copy leaked, and people tend to be more careful when they know copies are traceable.
- It does not take content down. If you need a takedown after finding the leaker, that is a separate step (a DMCA service or the platform's own report form).
FAQ
Can an invisible watermark really identify who leaked a screenshot? Often, yes, if the mark survived. Markedfile is designed to survive screenshots and recompression (except images under 256×256 px, which use a more fragile mark), but detection is heuristic and there is no 100% guarantee.
Do I need a different file for every person? Yes, that is the point: one marked copy per recipient. Markedfile creates them for you from the same original.
Will people notice the watermark? It is invisible in normal viewing. Look closely at both versions and the difference is not meant to be visible.
Does it work on PDFs too? Yes. Markedfile marks images and PDFs.
Is it free? There is a Free plan: 5 active files, detection for 6 months. Paid plans start at $9.99/month.
What if my photo already leaked? Markedfile cannot trace it, because that copy was never marked. Use it for the next time you share.
Mark each copy before you share it
Create a free account. You can mark 5 files and see whose copy leaked.