Redaction vs pseudonymization

All articles

Detection

1 min read

Redaction vs pseudonymization

The difference between removing data and replacing it consistently.

How it works

Redact

The applicant holds CPR .

The content stream is gone. A black bar is the visible result, not a cover you can lift.

Pseudonymize

The applicant Person 1 holds CPR National ID 1.

Tokens stay linked: Person 1, Email 1, and Address 1 belong to the same dossier.

Redact removes the content. Pseudonymize replaces it with a stable token so the document still reads.

Redaction

Redaction permanently removes the underlying content from the PDF's content stream. The data is gone, not just hidden behind a black box. When a box covers image content (a scan, photo, or embedded bitmap), that page is rasterized so the original pixels cannot be pulled out. Scripts, comments, and form fields are stripped.

How it works

Output scan

Page 2 still contains 010203-1234

Verification re-extracts text and leftover bytes.

Download refused

Sensitive text survived apply

Jump to the page, box it, finalize again.

Every export is verified fail-closed: if flagged text is still readable as page text or leftover file bytes anywhere you did not deliberately keep it, the export is blocked.

Verification reports where surviving text sits, so the document lists each location by page. Jump to it, drop a redaction box on it, then finalize again. Text left standing inside a box you skipped on purpose is allowed, and the redaction certificate states those keeps.

Pseudonymization

Pseudonymization replaces identifiers with consistent placeholders (for example Person 1, Email 2) across an entire case, so documents stay readable while removing identity. Tokens are identity-linked: Person 1, Email 1, and Address 1 belong to the same dossier.

Still need a hand?

Help center