PDF privacy
Deleted from a PDF, but still inside it
When a PDF is edited, many programs do not rewrite the file. They add the changes to the end and leave the earlier version where it was. A viewer shows only the latest one — but whatever you removed can still be read by anyone who looks at the file itself.
1 How a PDF gets saved
A PDF is a set of numbered objects — pages, fonts, images, text — and a table at the end saying where each one starts. The format lets a program save changes as an incremental update: it appends the objects that changed, and a new table that points back to the old one. Nothing before it is touched.
It is fast, and sometimes required: a digitally signed PDF can only be changed this way without breaking the signature, which covers the bytes that came before. Filling in a form, adding a signature or a comment, or a plain Save in many editors all append rather than rewrite.
So a line of text you deleted, a figure you changed, or the old title may all still be in the file, one revision back.
2 Check a PDF
Open it in hexscope — it reads the file in your browser and uploads nothing. A PDF that was changed after it was first saved says so: Edited: once after it was first saved; the earlier version is still inside. The structure is shown revision by revision, so you can see exactly which objects each update replaced, and the bytes of the old ones.
The sample is a one-page plan whose salary line was replaced in an update. The page you would see shows the new line; the first revision still holds the old page, and hexscope shows the line the update took off it: Taken off the page: Salary: 4,200 EUR a month.
3 What else a PDF says about you
- The author — often simply the name the writing program was set up with.
- The program it was written in, and the one that turned it into a PDF, with their versions.
- When it was created and last changed, with the time zone.
- An editing history, in the XMP metadata some programs add.
- A file identifier that stays the same across copies and edits, which can link a copy back to the original.
hexscope also marks what a PDF can run or carry: JavaScript, actions that start other programs, and attached files.
4 Get a copy without the earlier versions
Under What this document reveals, press Remove it — save a clean copy. hexscope writes the PDF anew, as a single version, with only what its pages use, each part copied byte for byte. The earlier versions, the author, the programs, the dates, the XMP metadata and the file identifier are left behind. Open the clean copy to check it.
In other programs, Save As to a new file usually rewrites a PDF from scratch, and printing it to a new PDF certainly does — at the cost of links and form fields.
Encrypted PDFs cannot be rewritten by hexscope yet.
5 A black box is not a redaction
Drawing a black rectangle over text hides it on screen and on paper, but the text is still there: selecting it, copying it, or searching for it brings it back. A redaction tool removes the text itself. Some tools only mark an area for redaction and remove it when you apply the redaction — until then, everything under the mark is still in the file.
hexscope walks each page the way a viewer paints it and finds text that a dark box was drawn over afterwards, and marks that were never applied, and text drawn so that no one can see it: invisibly, in white on white, off the page. It shows the text it found, so you can see what anyone else could read. Its clean copy takes that text out — letter by letter, leaving the rest of each line where it was — and applies the marks, as a redaction tool would. Open the clean copy to check it.
Check a PDF now Your files never leave your browser.
Incremental updates are part of the PDF standard, ISO 32000-1, section 7.5.6.