How to Redact a PDF: Free, Acrobat & Mac Methods (2026)

To redact a PDF, open it in a tool with a real redaction feature, mark the sensitive text (or let AI find it), apply the redaction, then save a flattened copy and check that the text under the marks can no longer be selected. The one rule that matters: drawing a black rectangle in a regular PDF editor is not redaction. The text stays in the file under the shape, and anyone can select, copy or extract it. That single mistake is behind nearly every "redacted" document leak.

How to redact a PDF in 5 steps

1. Open the PDF in a redaction tool. Not a viewer, not an image editor. The tool must be able to delete content, not just draw on top of it.
2. Find every sensitive item. Names, addresses, Social Security numbers, account numbers, dates of birth, signatures. AI redactors scan the whole file for these automatically; in Acrobat you search or mark them by hand.
3. Apply true redaction marks. The tool should remove the underlying text and image data where each mark sits.
4. Check every page, including headers, footers and metadata. Identifiers hide in page headers, repeated signature blocks, form fields, comments and the document properties.
5. Save a flattened copy and test it. Open the result, try to select the text under a mark and run a search for one of the redacted words. If either works, start again with a different tool.

Which method should you use?

Method 1: Redact a PDF for free with AI

A browser-based AI redactor such as Redact PDF AI reads the whole document in one pass, flags names, addresses, phone numbers, emails, IBANs, card numbers, dates and ID numbers in more than 100 languages, and exports a rasterised file. Scanned PDFs are handled by OCR first, so a photographed bank statement works the same way as a digital one.

1. Upload the PDF (or a JPG or PNG). The first document is free and needs no account.
2. Review the suggested marks. Remove any you want to keep visible, and add your own by dragging over text or images.
3. Add standing rules if you redact often: terms that are always removed (a client name, a case number) and terms that always stay (your company name).
4. Download the flattened copy. The original is deleted after processing and is never used to train AI models.

This is the fastest route when a document has dozens of identifiers, when the PDF is a scan, or when you have a stack of files.

Method 2: Redact a PDF in Adobe Acrobat Pro

Acrobat Pro (not the free Reader) has a proper redaction workflow. Reader can draw shapes but cannot remove content.

1. Open the PDF and choose Tools → Redact (in the new interface: All tools → Redact a PDF).
2. Mark text by dragging over it, or use Find Text to mark every occurrence of a word or a pattern such as phone numbers.
3. Click Apply (or Save). Acrobat warns that the change is permanent. This is the step that actually deletes the content; marks that are not applied remain reversible.
4. Run Sanitize document (Remove Hidden Information) to strip metadata, comments, bookmarks, form data and attachments that may still contain the original text.
5. Save under a new file name so the original stays intact.

Acrobat is reliable for one-off documents, but every item is selected by hand and the subscription is paid. Our Redact PDF AI vs Adobe Acrobat comparison covers when each makes sense.

Method 3: Redact a PDF in Preview on a Mac

Preview on macOS Big Sur and later has a real Redact tool in the Markup toolbar. The black rectangle from the Shapes tool is not redaction.

1. Open the PDF in Preview and click the Markup toolbar button (the pen tip).
2. Choose the Redact tool (the icon with the crossed-out text).
3. Drag over the text or image to remove. Preview deletes it immediately.
4. Save with File → Export as PDF under a new name. Preview flattens the redacted areas on save.

Preview cannot search for text, so it suits short documents only. The full walkthrough is in How to redact a PDF in Preview on a Mac. On an iPhone or iPad, Markup has no redaction tool at all; use a browser-based redactor in Safari as shown in How to redact a PDF on an iPhone.

Method 4: Redact a Word document or a scanned PDF

Word, Google Docs and Pages have no redaction feature. Black highlighting can be removed with one click, and a shape placed over text leaves the text in place. Either replace the sensitive words with "[redacted]" and export to PDF, or export to PDF first and redact the PDF with one of the methods above. Turn off Track Changes before you export; the revision history contains the original text.

Scanned PDFs are images without a text layer, so a text-based redactor cannot find anything in them. Use a tool with OCR: it reads the text inside the image, detects the identifiers and then rasterises the page on export. Without OCR you must draw every mark by hand.

What to redact in a PDF

The list depends on who will receive the file, but these items are almost always in scope:

• Names and addresses of people other than the recipient
• Social Security, passport, driver's licence and tax ID numbers
• Bank account numbers, IBANs, card numbers, routing numbers
• Dates of birth, phone numbers, email addresses
• Health information, religion, union membership and other special categories
• Signatures and faces in images
• Case numbers, client IDs and internal references when the document leaves the organisation

For a complete list with examples, read What is PII. The step-by-step guides for individual data types cover names, email addresses, phone numbers, addresses, IBANs and bank accounts, credit card numbers and dates of birth.

How to check that a PDF was really redacted

Three checks take ten seconds and catch almost every failed redaction:

1. Select test. Click and drag across a redacted area. If text highlights, the content is still there.
2. Search test. Press Ctrl+F or Cmd+F and search for a word you redacted. Any hit means the text layer survived.
3. Properties test. Open the document properties and look at the title, author, subject and keywords. Metadata often repeats the names you just removed.

If you need to go deeper, copy the whole page into a text editor or run the file through a text-extraction tool. How to unredact a PDF, and when you can't explains what recovery looks like from the other side.

Why a black box is not redaction

A PDF stores text, images and drawn shapes in separate layers. A rectangle drawn in a viewer is added as a new shape on top of the text layer; the text itself is untouched and stays searchable, selectable and extractable. True redaction does three things a shape never will:

• Removes the content under the mark from the text and image layers, usually by rasterising the page.
• Strips hidden data: metadata, comments, form fields, bookmarks, attachments and the revision history.
• Produces output that cannot be undone by copying, searching or analysing the file.

Court filings, law firms and large companies have all leaked data by shipping drawn boxes instead of applied redactions. The habit to build: always test the output before it leaves your machine.

Frequently asked questions

Can redacted PDF text be recovered? Only if it was never truly redacted. A drawn black box leaves the text fully recoverable by copy-paste or extraction. A properly applied, flattened redaction removes the data and cannot be reversed.

How do I redact a PDF for free? Upload it to a browser-based AI redactor such as Redact PDF AI, review the suggested marks and download the flattened copy. The first document is free and needs no sign-up. On a Mac, Preview's Redact tool is also free.

Can I redact a PDF in Adobe Reader? No. Adobe Acrobat Reader can draw shapes but cannot remove content. You need Acrobat Pro, Preview on a Mac or a dedicated redaction tool.

Is it safe to redact a PDF online? Yes, if the service encrypts the upload, deletes the file after processing and does not train AI on your documents. Redact PDF AI processes files in the EU and Switzerland, deletes them after download and never uses them for model training.

What is the difference between hiding and redacting text in a PDF? Hiding places a visual layer over the text while the original stays in the file. Redacting removes the underlying data so it cannot be recovered.

Can I redact a scanned PDF? Yes. Tools with OCR read the text inside the scanned image, detect identifiers and rasterise the result, so scanned and handwritten documents can be redacted like any other PDF.

How do I redact a PDF on a Mac or an iPhone? On a Mac, use Preview's Redact tool (macOS Big Sur or later) or a browser-based redactor. On an iPhone or iPad, Markup cannot redact; use a browser-based redactor in Safari. Both are covered in our Mac and iPhone guides.

Does redacting a PDF remove the metadata? Not automatically in every tool. Acrobat needs the separate Sanitize step; Redact PDF AI strips metadata as part of the flattened export. Always check the document properties after redacting.

Get started

Redact your first PDF free at redact-pdf.ai/redact-pdf, or learn what counts as PII before you start.

---

Need the reader to follow who is who? Anonymize a PDF with placeholders instead of black boxes · How to anonymize a document before ChatGPT