Every year, law firms, government agencies, and hospitals release documents they believe are redacted — only to have the underlying text extracted by anyone with a PDF reader and basic curiosity. The 2009 ACLU case that revealed NSA surveillance details, the Pentagon documents released with names still recoverable beneath black boxes, the court filings where attorney-client communications survived under a rectangle of black fill — all of these were the result of the same mistake: using a drawing tool to cover text rather than using a proper redaction tool to remove it. When you draw a black rectangle over text in Adobe Reader's drawing tools, Photoshop, or most online PDF editors, the text remains in the PDF's content stream. Any user can open the file in a text editor or use copy-paste to recover the hidden information. This is not an edge case or a technical exploit — it is the expected behavior of a visual cover-up that never actually modifies the underlying data.
LuraPDF performs true redaction: the regions you mark are processed so the underlying text characters, image pixel data, and content stream instructions are removed from the PDF structure before the file is written to disk. What replaces them is a solid filled rectangle — not a rectangle sitting on top of recoverable data, but a rectangle in place of data that no longer exists. Metadata — including author names, creation software, embedded document properties, and revision history that could reveal who created the document — is also stripped during redaction. The output PDF passes standard redaction verification checks. All of this runs locally in your browser with no upload. Your document never leaves your device, which is especially important when the document contains exactly the kind of sensitive information that makes proper redaction necessary in the first place.
Proper redaction is not just good practice — in many fields it is a legal requirement. Cover-up tools create liability. True content removal does not.
Redact client Social Security numbers, home addresses, and minor children's names from court filings before submission. Federal rules require actual content removal — a visual cover-up is a violation that can result in sanctions and mandatory refiling.
Remove SSNs, salary figures, disciplinary records, and medical information from HR documents before sharing redacted versions with auditors, counsel, or third parties — ensuring only the explicitly approved content is accessible.
Strip patient names, dates of birth, insurance numbers, and diagnosis codes from medical record PDFs before sharing for research, billing review, or inter-department transfer — with confidence that the PHI is genuinely absent from the file.
Redact account numbers, routing numbers, and transaction details from financial statements before sharing with partners, regulators, or clients who should only see aggregate or summary data.
Apply FOIA exemption redactions to responsive documents with true content removal so the released versions pass forensic verification and do not expose the exempt information through simple copy-paste.
Remove your SSN, date of birth, financial account numbers, and signature from scanned documents before emailing to service providers, landlords, or any party who only needs partial information from a form.
The technical difference between a cover-up and true redaction is invisible to the eye but decisive in a courtroom, an audit, or a data breach investigation.
The critical difference between cover-up and true redaction is at the content stream level. A standard PDF page has a content stream — a sequence of instructions that tell a renderer what to draw. Text instructions include the character codes, positions, and font references. When LuraPDF processes a redaction, it parses the content stream for each page, identifies the content instructions whose rendered output falls within the redaction box coordinates, removes those instructions from the stream, and writes a filled black rectangle in their place. The result is a content stream where the sensitive data instructions have been deleted, replaced by a solid fill. There are no invisible text characters under the rectangle — the rectangle is all that is in that region of the stream.
For image-based redaction, the pixel data within the redaction region of embedded raster images is overwritten with black pixels before the image stream is re-encoded into the PDF. For metadata, pdf-lib clears the document information dictionary fields — Title, Author, Subject, Creator, Producer — and removes the XMP metadata stream that modern PDF creators embed. Pattern redaction uses text extraction to find all matching strings across pages, calculates their bounding boxes from the text position data, and automatically adds those bounding boxes to the redaction queue before processing. Everything runs in your browser's JavaScript engine. The File API delivers the input PDF to memory, the redaction logic modifies content streams in place, and the Blob download API delivers the sanitized output to your device — no network request carries any part of your document.
| Feature | LuraPDF | Sejda | Adobe Acrobat |
|---|---|---|---|
| Runs in browser (no upload) | Yes | No | No |
| True content removal (not cover-up) | Yes | Yes | Yes |
| Automatic metadata stripping | Yes | Partial | Yes |
| Free with pattern redaction | Yes | Limited | Paid only |
Redaction is irreversible on the output file. A few verification steps before downloading protect you from missing a sensitive instance.
Always keep an unredacted backup of the original file before applying redaction — the operation is permanent on the exported document and cannot be undone
Use pattern redaction for repeating data like SSNs, email addresses, and account numbers — manually drawn boxes can miss instances on pages you did not review individually
After downloading the redacted PDF, open it in a separate PDF reader and attempt to copy-paste text from the redacted regions — if the paste returns blank, the redaction is genuine
Enable metadata stripping by default — even a visually clean redacted document can reveal the author, creation date, and software version through document properties
Never use a drawing tool, highlighting tool, or fill annotation to cover sensitive text and call it redaction — the underlying text remains in the file and can be extracted with trivial effort
For documents that will be filed with courts or regulators, verify against your jurisdiction's specific redaction standards — LuraPDF performs true content removal but jurisdictional compliance requirements vary
Your document never leaves your browser. Underlying text removed, metadata stripped, black boxes in place of data that no longer exists. True redaction — free, instant, private.