Check at least two data layers before sharing a photo or screenshot

The original is your only reliable fallback. Copy it first, then give the working file a name without a person's name, phone number, or order number. Make separate copies for separate recipients instead of repeatedly overwriting one file.

Remove areas unrelated to the conversation before redacting. Cropping reduces the surface to protect, but some editors save only a crop box and retain original pixels. Exporting a new flattened file is safer.

Names, ID numbers, QR codes, addresses, plates, chat avatars, and notification bars can identify someone. Blur and pixelation preserve structure and are risky for short text, patterned numbers, and high-contrast QR codes. Opaque blocks are easier to verify.

Extend the cover slightly beyond the text. Check reflections, thumbnails, message history, and interface autocomplete. For highly sensitive IDs or medical records, the safest option is usually not to use a public link or ordinary group chat.

Phone photos may include capture time, device model, orientation, and sometimes GPS. Screenshots often carry less metadata, but not every platform or editor cleans it. Treat metadata review as its own step.

Re-encoding visible pixels as a new JPG, PNG, or WebP usually drops many attached fields. It does not hide a visible street number, face, or chat message. Handle both layers.

Do not rely on the editor's layer preview. After export, close it and open the final file in a gallery or browser. Zoom into cover edges, then check filename, dimensions, and metadata summary. Confirm that the file being sent is the new one, not an original auto-selected by the chat app.

If a platform generates previews or keeps an original-download link, test the share method with a file containing no real data. If you cannot confirm the audience, do not treat “unlisted” as encryption or access control.