How to Redact a PDF So the Text Is Actually Gone
Drawing a black box over text does not remove it — the words are still selectable underneath. Here is what real redaction does and how to verify yours worked.
Updated August 11, 20265 min read
There is a long, embarrassing history of organisations publishing documents with black rectangles over the sensitive parts, only for a reader to select the text underneath and copy it out. It has happened to law firms, government departments and technology companies. The mistake is always the same and always understandable: the rectangle looks like removal, and the difference between covering and deleting is invisible on screen.
Why a black box is not redaction
A PDF is a set of drawing instructions layered in order. Text is one instruction — "draw these characters at this position with this font." A black rectangle drawn over it is another instruction, executed later, so it appears on top. Neither instruction knows about the other. The text instruction is still in the file, unchanged.
So the reader draws the text, then draws the rectangle on top, and you see black. But select across the area and you select the text. Copy it and you get the words. Run any text-extraction tool over the page and the redacted content comes out in full. Nothing has been removed.
What real redaction does
Proper redaction deletes the content objects themselves and then draws the black mark. Three things have to happen, and a tool that skips any of them is not redacting:
- 1The text objects covering that region are removed from the page's content stream, not merely obscured. After redaction there is nothing there to select.
- 2Any image data within the region is deleted or replaced. Covering part of a photograph leaves the pixels in the file if only a shape is drawn over them.
- 3The document is rewritten rather than incrementally updated. PDFs support incremental saves, where changes are appended and the previous version is still present in the file. An appended redaction can be undone by reading the earlier revision.
That third point catches even careful people. A file that has been edited and saved several times can carry its entire history. Rewriting the document flat is what discards it.
How to redact a document
Work on a copy
Redaction is destructive by design. Keep the original somewhere separate — if you redact the wrong paragraph, there is no undo once the content is gone.
Mark every region
Open the document in the PDF editor (redaction needs Pro) and draw redaction areas over each piece of sensitive content. Work through the whole document before downloading, not page by page.
Download the redacted file
The editor removes the marked content when it writes the file. Until you download, the marks exist only on screen.
Verify the result
Do not assume. Test it — the next section explains how.
Clean the metadata
Redaction handles page content. The metadata editor handles the title, author, keywords and creation history that travel alongside it.
Redaction is a Pro feature of the PDF editor: it removes the underlying content and rewrites the document, entirely on your device — nothing is sent to our servers.
Redact a PDFVerifying that it worked
Three checks, in increasing order of thoroughness. Do at least the first two on anything that matters.
- Select and copy. Drag across the black region in a PDF reader and press copy. If anything lands on your clipboard, the redaction failed.
- Search the document. Use the reader's find function to search for a word you redacted. Zero results is the answer you want.
- Extract all text. Run the file through PDF to text and read the output. This catches content that is present but not visually selectable, which the first two checks can miss.
For genuinely high-stakes documents there is a fourth option: after redacting, convert the pages to images and rebuild the PDF from those with images to PDF. This flattens everything, guaranteeing no text layer survives. The cost is a document that is no longer searchable and is usually larger, so reserve it for cases where the consequence of a leak outweighs both.
What else leaks besides the text
Redacting the visible content is necessary but not sufficient. Documents carry a surprising amount of context around the edges:
| Where it hides | What it reveals |
|---|---|
| Document metadata | Author name, organisation, software, creation and edit dates |
| Filename | Client names and case numbers are common in filenames |
| Comments and annotations | Internal review discussion, often unredacted |
| Form field values | Data entered into fields, retained even when not displayed |
| Attachments | Whole files embedded inside the PDF |
| Bookmarks | Section titles that describe what was redacted |
The bookmark case is a neat illustration of the whole problem: you redact a name from page 12, and the bookmark panel still reads "Correspondence with [name]." Nothing was covered incorrectly — the leak was somewhere nobody thought to look.
The same principle applies outside PDFs. Photographs carry GPS coordinates and device identifiers in their metadata, which is a different mechanism with the same consequence — see removing EXIF metadata from photos. And when the sensitive content is a face rather than a name, the face blur tool can hide it with a mosaic or a solid block. Both destroy the pixels rather than drawing over them. A plain blur can be partly reversed, so don't rely on it for anonymity.
Frequently asked questions
Why can people still read text under a black box in a PDF?
Because the box is drawn on top of the text rather than replacing it. The characters are still stored in the file and can be selected, copied or extracted by any text-extraction tool. Real redaction deletes the underlying objects before drawing the mark.
How do I check whether my redaction actually worked?
Select across the black area and try to copy — nothing should land on your clipboard. Then search the document for a redacted word and confirm zero results. For a stricter check, extract all text from the file and read the output.
Does flattening a PDF to images guarantee redaction?
It does remove any text layer, which is why it is used for high-stakes documents. The trade-offs are that the document is no longer searchable and the file usually gets larger, so it is worth reserving for cases where those costs are acceptable.
Is it safe to redact confidential PDFs with an online tool?
With ours, yes — redaction is a Pro feature of our PDF editor, and the document is opened, edited and rewritten on your own device, and never transmitted. For redaction specifically that matters more than for most tasks, since uploading the unredacted file to a server is precisely the disclosure you are trying to prevent.