Redact a PDF: remove names, phone numbers and IBANs for good

How it works

Tap or click the box to select your PDF, then type the words to redact, separated by commas. A typical case: you need to share a court document, a lease or a customer email thread, and a person's name, phone number and bank details must not be readable. Type "John Smith, (555) 010-4477" and set the ready-made finders to emails, phone numbers and IBANs. When you press "Redact", the tool scans the text of every page, deletes the matching letters from inside the document and puts a black box in their place. Case and spaces do not matter: typing "john smith" also finds "JOHN SMITH". The rest of the line stays where it was. The result line shows how many matches were removed on each page; if nothing is found, the file is returned unchanged. After downloading, open the file and try selecting text where the box is, or search with Ctrl+F: the removed text is gone. Text inside scanned pages is part of a picture and cannot be found by this tool.

This tool is also known as redact pdf, redact pdf free, black out text in pdf, remove text from pdf permanently, pdf redaction tool, hide personal information in pdf.

What is Redaction?

Redaction means removing confidential information from a document so that it cannot be recovered. Proper redaction does two things together: the text of the information is deleted from the file, and a black box shows the reader that something was taken out. Drawing a black rectangle alone is not redaction, because the text underneath is still in the file and can be copied out. When privacy rules such as the GDPR require personal data to be removed before a document is shared, only redaction in this full sense does the job.

What is PDF text layer?

A PDF has a text layer when its letters are stored as character codes rather than as a picture. PDFs exported from Word, Excel or a web page have one: you can select, search and copy the text because it is on the page letter by letter. This tool looks for the words to redact in that layer, deletes the letters it finds and draws a black box in their place. Pages scanned on a copier or photographed with a phone have no text layer; the words are just part of an image, so the tool finds nothing there and says so.

What is the difference between real redaction and drawing a black box?

Drawing a black box (a shape in Word, a rectangle comment in a PDF reader, or a black highlight) only covers the view: the text underneath stays in the file. Selecting the area and copying it, searching with Ctrl+F or deleting the comment reveals the information, and several published court filings and government reports have leaked data exactly this way. Real redaction removes the text from inside the document; the box only marks the gap. This tool does the second, and it also drops the old page content from the file so the text does not linger as an orphan.

Why a black box is not enough

A PDF page is built in layers: the text is on one layer, and a shape or comment added later sits on top of it. A black box on top hides the words on screen and on paper, but the letters below are still in the file. Selecting text where the box is and pasting it into an email, converting the file to Word or typing the name into the search bar brings the information back, and most PDF readers delete a comment with one click.

This tool removes the letters from the text layer. Each deleted letter is replaced by a gap of exactly the same width, so the rest of the line stays in place, and a black box is drawn over the gap. When it is done, the tool reads the result again with an independent PDF engine; if a match can still be extracted, that page is turned into an image and blacked out, and the result line says so. The redacted text is therefore neither extractable nor present in the file's bytes.

What gets redacted, and what does not?

Redacted: every match in the page text (headings, footnotes, table cells and text inside embedded graphics included), values typed into form fields and their appearance, the text of comments and notes, bookmarks and the title, author and subject in the document properties. If a term appears in the XMP metadata, that whole packet is deleted. The phone finder catches numbers of 9 to 15 digits that start with +, 0 or a bracket, or the 555-123-4567 pattern, and leaves dates and amounts alone; the IBAN finder only takes IBANs whose check digits are valid.

Not redacted: text inside scanned or photographed pages (it is a picture), files attached to the PDF and signature images. With some old fonts that draw letters as custom shapes instead of standard characters, the tool converts that page to an image; text on that page is then not selectable. A digital signature on the PDF becomes invalid, because the document has changed. The tool only finds what you type, so add ID numbers such as a Social Security number yourself.

Frequently asked questions

How do I redact text in a PDF for free?

Select the file on this page, type the word and press "Redact"; every occurrence is deleted from the document and covered with a black box. In Acrobat the same job is under Tools › Redact, which needs the paid version; drawing a rectangle in a free reader only covers the text and does not remove it.

Can redacted text be recovered?

No. The letters are deleted from inside the document and the old page content is not kept in the file, so copying, searching, text extraction or converting to Word will not return it. Keep your original, because undoing a redaction is only possible from the unredacted source.

Can I redact a scanned PDF?

Not with this tool: in a scanned page the words are part of a picture and cannot be searched as letters. The tool notices this, leaves the file unchanged and says it has no text layer. Run the scan through OCR first, or cover the information on paper and scan again.

Is the information I want to redact sent anywhere?

No. Searching, deleting and rewriting the file all happen in your browser; neither the document nor the names and numbers you type leave your device. Uploading a file to a server in order to protect personal data would defeat the purpose.