How to redact a PDF permanently — without Adobe Acrobat

When someone says "redact a PDF," most people reach for a shape tool and draw a black rectangle over the sensitive line. It looks correct. It is not. In a standard PDF, a drawn shape is an annotation layered on top of the original text. The text underneath is untouched.

The consequences are well-documented: government filings, court exhibits, and corporate memos have all been leaked because a reviewer could:

• Delete the black rectangle in any PDF editor.
• Copy the text from underneath the black bar and paste it into a document.
• Extract the underlying text stream programmatically with a few lines of code.
• Recover the author, revision history, and creation software from PDF metadata.

Permanent redaction has three requirements. Miss any one of them and the redaction leaks:

1. Destroy the pixel and text data under every black bar so nothing can be copied, OCR'd, or recovered from a lower layer.
2. Strip metadata — author, title, producer, creation date, modification history, and any XMP packet embedded by the source application.
3. Strip embedded content — attached files, JavaScript, embedded fonts that carry glyph tables, and image EXIF blocks that survived the export.

Adobe Acrobat Pro's redaction tool does the job, but it comes with real trade-offs: a paid subscription, a desktop install, and — critically — the original file lives on your machine (and possibly Adobe's cloud, depending on your settings) while you work. For a one-off redaction on a sensitive document, that is a lot of surface area.

You do not need Acrobat to redact a PDF permanently. Any tool that rasterizes the affected pages, paints opaque bars into the pixels, and rebuilds the PDF from that raster will achieve the same destructive result. The technique is what matters, not the vendor.

This is the approach Censr uses, and it is the same approach forensic teams recommend for hard-to-verify documents. Step by step:

1. Render each affected page to a bitmap at print resolution (typically 200 DPI). This flattens every text glyph, vector shape, and image into pixels.
2. Paint opaque black rectangles directly onto the bitmap, at the coordinates the user selected. There is no layer to peel back — the black is now part of the image.
3. Build a fresh PDF from the redacted bitmap. The new PDF has no text stream, no embedded fonts, and no history from the source document.
4. Scrub metadata — title, author, subject, keywords, producer, creator, creation and modification dates — all set to empty or epoch values.
5. Shred the original from memory once the redacted file is downloaded, so nothing lingers for recovery.

The output is visually identical to the source (minus the bars), but the underlying data model is completely new. There is nothing beneath the bar because there is no "beneath" — the bar is just pixels in an image.

Censr implements the rasterize-and-shred approach as a browser-only tool. Your PDF is never uploaded — the redaction happens locally in JavaScript and WebAssembly, and the original file is discarded from memory as soon as you download the redacted copy.

1. Open the Censr workspace. Drop in the PDF you want to redact.
2. Drag black bars over any text, signatures, faces, or numbers you want removed.
3. Download the redacted PDF. The bars are burned into the pixels; the metadata is stripped.
4. Close the tab. The original is gone from memory.

No account required to try it. There is a small fee per redacted download; the tool itself is free to use for as long as you want to test the workflow.

• Drawing shapes in a PDF viewer (Preview on macOS, most free editors) usually adds an annotation, not a permanent edit. Assume anything you draw is removable unless the tool explicitly says "redact."
• Printing to PDF flattens layers but preserves the underlying text stream in most modern printer drivers. Do not rely on it for redaction.
• Exporting to Word and back often preserves hidden text and comments in the round trip. Metadata is rarely stripped.
• Uploading to a redaction website means your unredacted document now lives on a server you don't control. Prefer tools that run in your browser.