Redact an image

Screenshots give away more than people notice: an email address in a header, a name in a chat, the last digits of a card, an order or tracking number, a QR code, or an open browser tab. Before you post one or attach it to a support request, drag a box over each thing that shouldn't be seen. Boxes can overlap, any single box can be removed with its own button, and Undo takes back the last one.

The style matters most for text. Blur and pixelation don't delete information, they average it, so each patch still depends on the letters underneath. With text in a known font, as in a screenshot, software can try candidate words, blur or pixelate them the same way and compare the results. Short strings such as card numbers have few enough possibilities for that to work. A solid black box keeps nothing of what was under it.

Blur and Pixelate are fine when the aim is to make something unrecognisable at a glance rather than impossible to recover: passers-by in a street photo, clutter on a desk, a house number in the background. They also look less stark than a black box. Both are applied heavily and only inside the box you draw, so nothing from outside it is mixed in. For several faces in one photo, the face blur tool finds them automatically.

Draw each box a little larger than the text, covering any letters that rise above or dip below the line, since a clipped edge can be enough to guess a character. This is safer than drawing a black shape in an editor that keeps layers, or covering text in a PDF, where the shape can sometimes be moved or deleted to reveal the original. Here the covered pixels themselves are replaced before the image is saved.