Redaction field manual
Privacy guides
Practical redaction guides for real sharing risks: fake redaction, EXIF leaks, PII scans and scanned PDFs.
Copy / select-all / inspect
Why black boxes on a PDF leak
Drawing a black box on a PDF looks safe, but the underlying text is still there and copyable. Here's how fake redaction works, real-world failures, and how to truly delete the text.
Strip GPS and camera data before you share photos
Phone photos embed GPS coordinates, capture time, and device model by default. Here's what that EXIF data reveals and how to remove it without losing quality.
The redaction self-check before you share a sensitive file
Before sending a contract, medical record, financial statement, or ID scan, run this checklist to avoid fake redaction and metadata leaks.
How to check whether a PDF is actually redacted
Three no-tool ways to verify a PDF's redaction held, plus a one-click re-checker. Text under the box = not redacted.
Adobe's Redact and Sanitize are two different things — miss one and you leak
In Acrobat, removing visible text (Redact) and clearing hidden info (Sanitize) are separate steps. Doing only one is the most common redaction failure.
Redact a screenshot before you share it — bug reports, order proofs, asking for help
Screenshots leak privacy you didn't notice: notification previews, browser tabs, account emails, shipping addresses. Mask locally before sharing — don't upload to a screenshot site.
Redaction by scenario: what to cover in contracts, medical records, and resumes
Different documents have different sensitive fields. This scenario-by-scenario checklist covers contracts, medical records, bank statements, and resumes — plus the corners people miss.
How Chinese national IDs and card numbers get detected in a document
Good PII detection isn't 'flag any 18-digit number.' National IDs use a mod-11-2 check digit and cards use Luhn — math separates real numbers from random strings.
Why scanned PDFs are harder to redact than digital ones
In a scan, text is pixels in an image, not selectable characters. That flips redaction and PII detection — the upside is it's harder to fail, the downside is detection needs OCR first.