Guides

How to redact a PDF properly

Redacting means removing information from a document before you share it: a name, an account number, an address, a whole paragraph. On paper, a thick black marker does the job. In a PDF it often does not, and redaction failures in court filings and official reports regularly make the news because someone drew a black box and assumed the text underneath had gone.

This guide explains why the usual shortcuts fail, what proper redaction does, and how to redact a PDF with A2ZDocs Redact without uploading it anywhere. It ends with the checks to run before the file leaves your hands.

Why black boxes, highlighter and shapes fail

A PDF page is built in layers. The text is stored as real characters, and anything you add on top, such as a black rectangle, a highlight or a shape, is just another layer drawn over it. Covering text hides it from your eyes, but the characters are still in the file. Anyone who receives it can get them back in seconds:

  • Search: press Ctrl+F (Cmd+F on a Mac), type a word you covered, and the viewer finds it under the box.
  • Copy and paste: select the covered area, copy it and paste it into a text editor. The hidden text appears.
  • Remove the cover: in many PDF editors the box is a separate object that can simply be moved or deleted.
  • Extract the text: tools that pull all the text out of a PDF ignore boxes completely.

Black highlighter in a PDF viewer is the same thing: a coloured mark over text that is still there. Making the text black on a black background, or covering it with a white box, fails in exactly the same way. In A2ZDocs, the White-out and shape tools in Edit PDF change how a page looks; they are not redaction.

There is a second, quieter leak: metadata. A PDF carries document properties such as the title, author, subject and keywords, often with a second copy in a format called XMP. A file whose title is a client's name, or whose author field holds your colleague's full name, gives information away even when every page is clean.

What real redaction does

Proper redaction removes the information from the file, not just from view. The text under each box is deleted from the page content, and a box is drawn where it used to be so readers can see that something was taken out. Links, comments and form fields in the area go too, and any picture that contained the information is replaced. Afterwards there is nothing to search for, copy or uncover.

What A2ZDocs Redact removes

When you save, Redact works through each page you marked:

  • Text under a box is deleted from the page itself. Every other character stays exactly where it was, so the rest of the page remains searchable and selectable.
  • Vector drawings that lie completely inside a box are removed.
  • Links, comments and form fields that touch a box are removed.
  • If a picture, a colour gradient or other content that cannot be cut out cleanly sits under a box, that whole page is turned into an image with the box burned in, so nothing hidden survives. Text on that page can then no longer be selected.
  • The old page content that held the removed text is dropped from the file, so it does not linger inside it.
  • The document properties (title, author and so on) and their XMP copy are removed as well, unless you untick that option.

For the strictest result, tick "Turn every redacted page into an image". Every page with a box then becomes a flat picture, at the quality you choose (Smaller, Standard or Sharp). Text on those pages can no longer be searched or copied, but nothing of the original page content remains. Pages without boxes are left as they were.

How to redact a PDF with A2ZDocs

  1. Open Redact and choose your PDF. It opens in your browser and is not uploaded. If the file has a password, remove it first with Unlock, using the password you already know.
  2. Drag a box over each thing you want to hide. The Redact area tool is already selected; switch to Select to move or resize a box.
  3. To find text automatically, type words or numbers under Find & mark, separated by commas, and select Mark matches. You can also tick Email addresses, Phone numbers or Credit card numbers to mark those wherever they appear.
  4. Go through every page and check each box. Automatic matching is a helper, not a guarantee: a name may be spelt differently, split across two lines, or appear inside a picture.
  5. To hide whole pages, select This page, or type page numbers such as 2, 5-7 and select Mark pages.
  6. Under Look, choose the box colour and, if you like, a label such as REDACTED to print inside each box.
  7. Leave "Also remove the document properties" ticked, choose any other options, and select Redact & save. Your new file ends in _redacted.pdf, and a short report says how much was removed.

Scanned documents

A scan is a picture of a page, so there is no text for Find & mark to search. You can still drag boxes over a scan by hand. Because the content under each box is a picture, the page is saved as a new image with the box burned in. If you want automatic matching, run OCR first to add a text layer, then redact. Redacted scanned pages are saved as images, so run OCR again afterwards if you need the rest of those pages to be searchable.

Check the result before you share it

Always test the redacted file, ideally in a different PDF viewer from the one you used to make it:

  1. Search for each word, name or number you redacted. The viewer should find nothing.
  2. Select the text around each box, copy it and paste it into a plain text editor. The redacted words should not appear.
  3. Run the file through PDF to text and search the .txt file it gives you. This shows all the text left in the PDF, including anything you cannot see on screen.
  4. Open the document properties in your viewer (often under File, then Properties) and check that the title, author, subject and keywords are empty or harmless.
  5. Look at the bookmarks and attachments panels. Redact does not touch bookmark titles or attached files: rename or delete bookmarks with Bookmarks and remove attached files with Attachments.
  6. Check the file name. A name like smith-medical-report.pdf undoes careful redaction.

Cleaning up metadata on its own

Redact removes the document properties by default. If you unticked that option, or want to tidy a file you are not redacting, use Edit metadata. It shows the current title, author, subject, keywords, the program the file was created with and the PDF producer. You can change any of them, or tick "Remove all metadata instead" to strip them all, including the XMP copy.

A short checklist

  • Never rely on black boxes, highlighter, white boxes or text colour to hide information.
  • Redact every occurrence, on every page, including headers, footers, tables and captions.
  • Remember pictures: signatures, stamps and photos of ID cards need boxes too.
  • Keep the unredacted original somewhere safe, and share only the _redacted copy.
  • Test the result with search, copy and paste, and PDF to text before you send it.

Because A2ZDocs works in your browser, the unredacted document stays on your device the whole time, which is exactly what you want when the reason for redacting is that the information is sensitive. The Privacy Policy explains how this works.

Tools in this guide

All guides