What is redaction? Definition and how it works

Redaction is the process of permanently removing sensitive information from a document before it is shared, so the protected details cannot be read, copied, or recovered. The black bars you see across a contract, court filing, or report are the visible sign of it, but proper redaction also deletes the text or image sitting underneath each bar.

That last part is what separates redaction from simply hiding something. If the underlying content is still in the file, the document has been covered, not redacted.

What redaction actually removes

A redaction targets specific pieces of information inside a document that is otherwise fine to release. The rest of the page stays readable. Common targets include names, home addresses, dates of birth, Social Security numbers, bank account and card numbers, phone numbers, email addresses, signatures, medical details, and anything a privacy law protects.

The goal is release, not destruction. A court can publish a filing once witness contact details are gone. A company can share a signed agreement once the pricing and personal data are stripped. A person can send a bank statement once the account number is removed. In each case the document keeps its value while the risky parts are taken out.

How redaction works, step by step

Digital redaction usually follows the same sequence, whether a person does it by hand or software helps:

1. Read the document and decide what has to go. Some of this is judgment (is this detail sensitive?) and some is pattern-based (this looks like a card number).
2. Mark each item, usually with a box over the text or image.
3. Remove the content under every box. This is the step people skip. The characters, image pixels, or vector shapes have to be deleted from the file, not just hidden.
4. Verify that nothing readable remains where a box sits, including hidden metadata and any copy of the text in the document's structure.
5. Save a flattened output so the boxes cannot be moved, deleted, or peeled back.

On paper the old method was a black marker and a photocopy. You marked the original, then copied it, and shared only the copy, because the copy has no lifted ink to recover. Digital files need the equivalent: the sensitive data has to leave the file itself.

Redaction versus covering text

Drawing a black rectangle in a PDF viewer or a word processor sits on top of the words. The text underneath is untouched. Anyone can select it, copy it, paste it somewhere else, and read every character. Search tools can find it. The file's internal text layer still holds it.

A quick test tells you which one you are looking at. Try to highlight and copy the text under a black bar. If it copies out, it was never redacted, only covered. Here is why that mistake keeps happening, and it is the reason redacted files sometimes leak the exact information they were meant to protect.

Types of redaction

Redaction shows up in a few forms depending on what you are removing and how you find it.

• Manual redaction, where a person reads the document and marks each item by hand. Slow but precise, and the default for short files.
• Pattern-based detection, where software scans for known formats like Social Security numbers or emails and boxes them automatically for review.
• Text, image, and metadata redaction, since sensitive data can live in the visible text, inside a scanned image, or in the file's hidden properties. A thorough job clears all three.

Where the word comes from

Redaction traces back to the Latin redigere, roughly "to bring back" or "reduce." In editorial use it once meant preparing and revising a text for publication. The privacy meaning is a narrower version of the same idea: preparing a document for release by taking the sensitive parts out. It is also why you will hear redaction called sanitizing a document. If you want the adjective side of this, see what redacted means.

FAQ

Is redaction the same as deleting the document?

No. Deleting removes the whole file. Redaction removes only the sensitive pieces and keeps the rest readable, so the document can still be shared.

Can redacted information be recovered?

It should not be, if the redaction was done properly. When content is only covered rather than removed, it can be recovered by copying the text or moving the box. When the underlying data is deleted from the file, there is nothing left to recover.

Is a black box the same as a redaction?

Only if the content under it is gone. A black box is the visible marker. Proper redaction is the removal of whatever was beneath it.

Redacting a file yourself

If you need to redact a PDF or image, use a tool that removes the content instead of covering it. Ours runs entirely in your browser, so the file is never uploaded, and what you take out is deleted from the file for good. Manual redaction is free and needs no account.

Redact a PDF for free.

Related reading: what does redacted mean, how to redact a PDF, is PDF redaction reversible.