Redact PII Before Sharing: Screenshots, Documents, Photos

Forwarding screenshots is a daily activity in modern work. Forwarding screenshots that contain personal information about colleagues, customers, friends, or yourself is a daily compliance risk. The checklist below takes about a minute and keeps the original in your browser the whole time.

The five-minute checklist

1. Pause before forwarding. Look at the entire image. Personal information hides in the corners: the browser tab next to the one you screenshotted, the OS clock, a notification preview, a sticky-note address.

2. Identify what needs to go. Walk through the image and tag every region:

• Names (yours, colleagues', customers')
• Email addresses, phone numbers
• Account numbers, transaction IDs, order numbers
• Addresses, GPS coordinates
• Faces (especially of people who haven't consented)
• Signatures, dates of birth
• Any external service login state visible in browser chrome
• File paths that reveal usernames (/Users/firstnamelastname/...)

3. Pick the redaction mode per region. Black bar for short critical things; blur for layout-preserving body text; brush for irregular shapes. (See Black Bar vs Blur vs Brush for the long version.)

4. Redact in Image Redactor. Drag, click, repeat. The live preview shows you exactly what the recipient will see.

5. Strip metadata. Even after redacting pixels, the file metadata can reveal you. A screenshot's EXIF on iOS includes the device model and timestamp. Run the redacted output through EXIF Remover.

6. Verify. Open the redacted file in a fresh viewer (Preview, Quick Look). Zoom in to each redacted region. Black bars should be solid; brush strokes should fully cover the underlying shape; blur should not preserve readable letterforms.

7. Delete or archive the original. A stale unredacted copy in ~/Downloads is the most common way redactions are undone: someone shares the wrong file by mistake.

Common landmines

The OS clock and battery percentage are part of every screenshot. Usually harmless, but together they pin you to a specific moment.

Browser tabs. A row of tabs across the top of a screenshot tells the world what services you use, sometimes including login-state hints.

Slack / Teams / Discord workspace names in the corner of any in-app screenshot. Sometimes those names are confidential.

File paths in error messages and stack traces routinely contain your OS username.

Email signatures in screenshotted inbox views.

Map embeds. A pin location is your home address.

Document scans

PDFs are a separate case. A PDF page contains text that can be selected and searched, and redacting visually in an image editor does not remove the underlying text layer. For PDFs with sensitive content:

• Use PDF Metadata Remover for author/title/creation metadata.
• For actual page content redaction, convert the page to an image first (PDF to Images), redact in Image Redactor, then optionally re-bundle (Images to PDF).

Photographs of people

Faces are PII in most jurisdictions when paired with a name or context. If you are publishing a photo where the subjects did not consent to be public:

• Brush mode in the redactor; trace the face outline.
• Or apply black bar across the eye region only (the photographer convention), survivable for most legal contexts, but check your local rules.

Generic group photos at conferences with badges visible: name redaction matters as much as face redaction.

The browser-only promise

Every step above runs in your browser. The redactor uses HTML5 Canvas; the EXIF remover walks the file bytes locally; the PDF tools use WASM. Open DevTools, switch to the Network tab, filter Fetch/XHR, and the row stays empty. The sensitive document never reaches a server.

That is the point. You are sharing the redacted version because you trust the recipient with it. The unredacted original stays with you.