Can redacted PDF text be recovered?
Short answer: with most tools, yes. A redaction is only safe if the underlying text and its geometry are actually removed and the result is verified. Many real-world leaks have come from "redacted" PDFs where the text was merely covered.
The three ways "redacted" text leaks
1. The text is still there
If the tool only drew a black box, the characters remain in the content stream. Select-all and copy, or a one-line extraction script, pulls them straight out.
2. The geometry gives it away
Even when characters are deleted, the spacing and positions of the remaining glyphs can encode the length and identity of what was removed — the glyph-position leak documented in Story Beyond the Eye (arXiv:2206.02285).
3. Metadata and old revisions
PDFs keep document metadata, embedded thumbnails, and incremental-update history. A pre-redaction copy can survive in any of those even when the visible page looks clean.
How to know your redaction is safe
The only reliable test is to attack your own output: try to extract text in the redacted region, inspect the glyph positions, scan the metadata, and check for retained revisions. If nothing comes back, it's safe. SecureRedact runs exactly this audit automatically and refuses to hand you a file that still leaks. Redact and verify a PDF →