How to check a PDF for hidden information
- Drop the PDF. It is read here in the browser; nothing is sent anywhere.
- Read the report. Each finding says what it is, why it matters, and shows you the actual values.
- Choose what to remove, and download. The switches start on whatever was found.
What a PDF carries that you can't see
A PDF is a container, and page content is only part of what goes in it. Export one from a word processor and it will typically record your account name, the application and version, and timestamps. Send a document round for review and every comment keeps the name of whoever wrote it, long after the comments themselves have been resolved. Fill in a form and the answers are stored as separate data next to the page, not as marks on it.
None of that appears when you open the file, which is why it gets sent. The failures that make the news are usually this, rather than anything sophisticated: a document that named an author who was supposed to be anonymous, or a review copy that still had the internal comments in it.
What the scan looks for
- Document information: author, creator, producer, and the created and modified dates.
- XMP metadata: a second, richer copy that many strippers miss.
- JavaScript and automatic actions: code set to run when the file opens or a page is shown.
- Attached files: whole documents carried inside the PDF.
- Comment authors: names attached to notes and mark-ups.
- Form data: values sitting beside the page rather than on it.
- Outbound links: where the document points, including tracking parameters.
- Layers: content that is switched off rather than removed.
- Signature fields, flagged because editing will break them.
What it doesn't claim
It cannot tell you whether something on the page should have been removed. Text under a black rectangle is normal page content and looks like every other piece of text in the file. If material must be unrecoverable, redact it properly, which rebuilds the affected pages from pixels so the hidden content stops existing.
For the narrower job of clearing document information only, the metadata remover and anonymiser do that in one click. To lock a finished document down, see protect PDF.
Common uses for the privacy scanner
- Check before you send: see whose name is on a document and what software made it.
- Clear a review round: strip comment authors from a document that went out for feedback.
- Publish anonymously: remove the metadata that would identify the author of a submission.
- Vet a file you were sent: find scripts, attachments and automatic actions before opening it in a reader.
Frequently asked questions
What does a PDF give away about me?
Usually your name and the software you used, from the document information the exporter wrote automatically. Often the dates you created and last changed it. Sometimes more: comment authors from a review round, the values in a form you filled in, files attached inside the document, scripts set to run when it opens, and layers that were switched off rather than removed.
Is this different from removing metadata?
Metadata is one of the things it looks for. The scan also covers comment authors, form data, attachments, scripts, automatic actions, layers and outbound links, and it tells you what it found rather than silently stripping everything.
Will cleaning change how the document looks?
No, with one exception you choose: flattening a form draws the values into the page, which is a visual change in the sense that the fields stop being editable. Everything else lives outside the page content and comes off without touching a pixel.
Can it tell me if something was badly redacted?
Not reliably, and it doesn't pretend to. Text hidden under a black rectangle is ordinary page content, indistinguishable from text that is meant to be there. If you need something gone for certain, rebuild those pages with the redaction tool.
Why can't it remove layers or signatures?
Both are structural. Removing a layer means deciding what to do with the content inside it, and silently invalidating a signature would defeat the reason it exists. The scan flags them so you can decide.
Is my document uploaded?
No. A tool that reads out everything hidden in your documents would be a strange thing to run on someone else's server. The scan and the cleaning both happen in this tab.