How to remove hidden text from a PDF (so it can't come back)
Drawing a black rectangle over text does not delete it. To truly remove hidden text you must strip it from the PDF's content stream and erase the glyph positions that can reconstruct it. Here's why, and how to do it safely.
Why "hidden" text in a PDF usually isn't gone
A PDF stores text as instructions in a content stream, separate from anything drawn on top. When a tool places a black box over a name or number, the box is just another drawing instruction — the original characters remain in the stream underneath. Anyone can select, copy, or extract them with a script or even a careful copy-paste.
It gets subtler. Even when a tool deletes the characters, it often leaves the surrounding glyph positions and advance widths intact. Those positions encode how wide the removed text was and where each character sat — enough to reconstruct the original string. The 2022 study Story Beyond the Eye (arXiv:2206.02285) showed most online redactors leak exactly this way.
What actually removing the text requires
- Delete the text operators for the sensitive region from the content stream.
- Remove or reflow the glyph positions so the geometry can't be used to infer what was there.
- Rasterize the region at print resolution as a belt-and-suspenders layer.
- Strip metadata and history — XMP/Info fields, embedded thumbnails, and previous saved revisions can each hold a copy.
- Verify — re-scan the output and confirm nothing is recoverable before trusting it.
Do it in your browser, free
SecureRedact does all of the above locally — your file never leaves your device — and runs a recovery audit on the result, blocking the download if anything is still extractable. Open the redactor →