Detect and redact PII in PDFs

Detect PII scans a PDF's text for personal data, Auto-Redact PII removes every match in one pass, and Verify Redaction re-scans the result to prove nothing extractable remains. The whole loop runs locally — the document never leaves your machine.

Detect PII

Privacy › Detect PII reads the extractable text of one PDF and reports every hit it finds, grouped per page — the result reads as "N items found across M pages". Previews are masked (e.g. jo********@om), so the scan output itself is safe to look at on a shared screen.

Six categories are matched:

The scan only sees text the PDF actually contains. A scanned document — a photo of a page — has no text layer, so the report comes back empty even though a human could read it. The app says so in that case: run OCR to see what the scan would find, or redact the visible regions by hand with Redact Content.

Auto-Redact PII

Privacy › Auto-Redact PII takes the scan and removes every hit in a single pass. It is a Pro feature.

Two things make this real redaction rather than a black box drawn on top:

• Affected pages are rasterized — rendered to an image with the redaction painted in — so the text underneath no longer exists in the file at all, not merely hidden under a rectangle.
• The pass fails closed: if a page cannot be rendered safely, the run reports the failure instead of shipping a cosmetic cover-up.

The trade-off: rasterized pages become images of text. They display and print identically, but that page's text is no longer selectable or searchable — for a PDF, that is the point.

Verify Redaction

Privacy › Verify Redaction runs the same scan on your output file and reports what is still extractable: the remaining text length plus any leftover PII hits. "Clean — no extractable text remains" is the result you want before a document leaves your hands. If it reports "N items still extractable", re-run the scan to see what survived — a pattern the matcher missed, or text on a page that wasn't touched.

Verify after every redaction. It is the difference between "looks blacked out" and "provably gone".

When to destroy instead

If the source document shouldn't exist at all once the clean copy ships, File Shredder overwrites the original with random data — three passes by default — before deleting it. That is permanent; there is no recovery.

Related

• Remove metadata from photos and PDFs
• Privacy and network behaviour
• Redact Content tool — manual, region-based redaction
• OCR tool — for scanned documents with no text layer
• Metadata hygiene guide