Data used for redaction should not become another upload record.

The site has no processing endpoint for images or text. File decoding, redaction, matching, scanning, and re-encoding happen in the browser.

Redaction layers, text, and results exist only in memory. The site cannot restore them after close or refresh, so save the files you need.

The image page aggregates start, success, download, and error events to measure usability. Counts contain no image, filename, rectangle, output, IP, or user identifier and are not sent when DNT or GPC is enabled.

Aggregated page counts without tool input may also be recorded. Ad scripts may use cookies, but ads appear only after the full explanation and away from upload and download controls.