Why "upload to redact" is a privacy problem
To redact a document is to hide its most sensitive parts. Uploading that same document to a third-party server to perform the redaction means the unredacted file, with everything still in it, travels across the network and sits on a remote server, however briefly. You are trusting a deletion policy you cannot verify.
Scrubbr removes that step. There is no upload endpoint. The redaction, the flattening, and the verification all run as JavaScript and WebAssembly on your machine.
Real redaction, then proof
- Draw boxes over names, account numbers, or any sensitive text.
- Marked pages are rasterized and rebuilt as images, so the underlying text and vectors no longer exist in the file.
- Scrubbr re-extracts text from the output and confirms none of the redacted content remains.
- Export a proof-of-redaction certificate (Pro) recording the file hash, timestamp, and verification result.
Frequently asked questions
Is the redacted text really gone?
Yes. Redacted pages are flattened to images and Scrubbr re-extracts text from the output to confirm none of the content under your boxes is recoverable by the parser.
Does my file leave my device?
No. There is no upload endpoint. Open your browser network tab and confirm no request carries your file.