Check a redacted PDF
Find out whether a PDF's blacked-out text is really gone, or still readable underneath.
Drop your file here
Processed securely and deleted within the hour.
The question this answers
Somebody sent you a document with black boxes on it. Are those boxes covering the text, or removing it? You cannot tell by looking — that is precisely the problem — and the difference is whether the information is public.
Drop the file in and this will tell you. It renders each page, finds the solid dark rectangles, and asks the PDF what text sits underneath each one. If anything comes back, the document is not redacted; it is decorated, and anyone can select the words with a mouse.
Use it on your own documents too
Before you send a redacted file, check it. Most PDF software offers a "black out" or highlight tool that draws a rectangle and does not remove anything, and it is very easy to believe you have redacted a document when you have annotated it. This takes a few seconds and answers the question definitively.
The other places information hides
Text under a box is the famous failure, and it is not the only one. This also reports on:
Embedded attachments and JavaScript — a PDF can carry whole files inside it, and scripts that run when it opens. Neither appears on any page.
Form fields — these store their value as text separately from what is drawn. A field displaying asterisks can hold the real value.
Annotations and comments — including the review discussion about the thing that was redacted, which is a genuinely common leak.
Page thumbnails — small pictures of the page, embedded for a preview pane. They are not regenerated when the page changes, so a thumbnail can show the page as it looked before the black box was added.
Metadata — author names, original filenames and the software used. Frequently more revealing than people expect.
Earlier versions of the file — this is the subtle one. A PDF saved incrementally appends changes rather than rewriting, so the document can still contain its own previous revisions, including the unredacted one. The check accounts for files linearised for fast web view, which legitimately look similar and are not a problem.
Searching for something specific
If you know what should have been removed, type it in. The file is decompressed — every stream expanded and object streams disabled so nothing can hide inside a compressed container — and the raw bytes are searched for that text in every encoding a PDF can store it in. This finds text that is in the file but on no page at all, which extraction-based checks miss entirely.
What a pass means, and what it does not
It means these particular checks found nothing recoverable. It is evidence, not a guarantee. This looks for text under solid dark rectangles; a redaction drawn in a pale colour, or as an image pasted on top, may not be recognised as a covered area. It cannot tell you whether the right things were redacted, or whether something sensitive was left in plain sight.
Nothing you upload is kept. The file is examined in a directory of its own and deleted as soon as the answer has been sent — and the examination is strictly read-only, which is less obvious than it sounds: an earlier version of this tool used a library call that quietly rewrote the file's metadata as a side effect of reading it, and then reported the metadata it had just added.
Common questions
How can text still be there if I cannot see it?
Because a black rectangle is drawn on top of the text rather than replacing it. The text remains in the file and any reader can select and copy it.
What should I do if it fails?
Do not send that file. Redact it properly — removing the content rather than covering it — and check it again.
It found no black boxes. Is that bad?
It means the document either has not been redacted by covering, or the covering is not a solid dark rectangle. It is not a failure, just a case this check cannot speak to.
Why does it warn about page thumbnails?
A thumbnail is a stored picture of the page, and it is not regenerated when the page changes. It can show what the page looked like before the black box was added.
What does "earlier versions of the file" mean?
PDFs can be saved incrementally, appending changes rather than rewriting. The unredacted original may still be inside the bytes. Files linearised for fast web view look similar and are not a problem, and are not flagged.
Does a pass mean the document is safe?
It means these checks found nothing recoverable. It cannot tell you whether the right things were redacted, or whether something sensitive was simply left visible.