On paper, a black marker works. Once the ink is on the page, the words under it are gone. That habit carries over to PDFs, where almost every editor offers a black rectangle or a black highlighter. The page looks exactly like a redacted printout. It isn't one.
This page explains what is actually inside a PDF, why a box on top doesn't remove anything, the ways hidden text leaks out, and how to tell whether a PDF you received or are about to send was redacted for real. For the full how-to, see how to redact a PDF properly.
What a PDF page is made of
A PDF page is not one flat image. It is a list of instructions: put this text here in this font, place this image there, draw this line. A viewer reads the list and paints the result on your screen. The text is stored as text, which is why you can select it, search it and copy it.
When you add a black rectangle, the editor adds one more instruction near the end of the list: paint a black shape at this spot. The viewer draws the text first and the box last, so the box hides the text from your eyes. Nothing in the list was deleted. The words are still in the file, in the same place, in the same font.
A black highlight works the same way. It is an annotation laid over the text, and annotations can be hidden, moved or deleted by anyone with a PDF editor.
How the hidden text gets out
None of this takes special skill. These are the everyday ways it happens:
- Copy and paste. Select the area, or the whole page, copy, and paste into a note. The covered words come along.
- Search. Ctrl+F or Cmd+F finds text under a box, and many viewers highlight the hit through it.
- Moving the box. In any editor that can edit annotations, the box can be dragged aside or deleted.
- Text extraction. Converting the PDF to Word or plain text pulls out every text object, covered or not.
- Screen readers and search engines. Software that reads the text layer never sees the box at all.
The NSA's guide Redacting with Confidence describes the same problem. Covering text with black rectangles, it says, "is a common and effective means of redaction for hardcopy printed materials. It is not effective, in general, for computer documents." It calls covering text with black "the most common mistake."
It isn't only text
The same guide warns about images. Covering part of a photo or a scanned signature with a separate black shape leaves the full image in the file underneath. Shrinking an image to make it unreadable doesn't work either, because the original pixels may still be there at full size.
Scanned documents have their own trap. A scan is a picture, but if it went through text recognition at some point, it also carries an invisible text layer behind the picture so you can search it. You can paint a box over the picture and still leave the recognized words in that hidden layer. The guide to OCR in plain English explains how that layer gets there.
Then there is information that never shows on the page: metadata such as the author and title, comments, bookmarks and form field values. The NSA guide lists metadata and document properties as the third common source of leaks.
How to tell whether a PDF was really redacted
You can check any PDF, whether you made it or received it, in about a minute:
| Test | What a real redaction shows | What a fake one shows |
|---|---|---|
| Search for a word under a box | No results | The word is found |
| Drag to select across a box | Nothing selects, or only the box | Text highlights under the black |
| Select all, copy, paste into a note | Covered words are missing | Covered words appear |
| Click the box in an editor | Can't be moved; it is part of the page | Moves or deletes like any shape |
If you only have the PDF open in a browser, the first three tests are enough.
What actually removes the content
There are two reliable approaches.
Delete what is under the mark. Proper redaction tools remove the text and image data inside the marked area when you apply the redaction. Adobe's Redact tool in Acrobat Pro works like this, and it offers to sanitize the file of hidden information as you apply. Apple's Preview on a Mac has a Redaction Selection tool that removes the selected text from view; Apple notes that it becomes permanent when you close the document, so redact a duplicate.
Flatten the page. Turn each page into an image with the black areas painted in, and build a new PDF from those images. There is no text layer left, so nothing can sit under a box. This is what the free redact PDF tool on this site does. It redraws every page at 150, 200 or 300 dpi, fills your marks solid black, and saves a new file from the images alone, without the original's metadata. The file never leaves your browser.
Flattening has a cost: the result is not searchable and is usually larger than the original. That is the trade for having nothing left to leak. If you don't own Acrobat, redacting a PDF without Acrobat walks through the free options.
Before you send anything
Most black-box mistakes happen on a deadline: a statement for a mortgage application, a pay stub for a landlord, a form for a school. Take the extra minute to run the search test on the file you are about to attach, not the one you edited. And if the document is going by email or a shared link, the guide to sharing documents securely from an iPhone covers the rest of the trip.