6068redact
Removes personal identifying information from PDFs before they go to an AI system. Supports Cal. Bus. & Prof. Code § 6068.1.
Loading redaction engine…
Stopped: pages without a text layer
These pages have no recognized text, so they cannot be checked for personal information. No output was produced.
Run Acrobat 'Recognize Text' with the 'Searchable Image' setting on the whole file, then upload it again.
Done
If no download started, this window may not support downloads (for example, an app's built-in preview pane). Open this page in Chrome, Edge or Safari and run it again.
Everything runs in this browser tab. Your file is never uploaded, and the app makes no network requests after it loads. You can disconnect from the network before choosing a file.
While you download it, the redacted output is kept in this browser's private storage on this computer. It is deleted when you start another run or reopen the app.
Built with MuPDF.js (AGPL-3.0, Artifex Software) and Arimo (SIL OFL 1.1). Source code (AGPL-3.0): github.com/iancorzine/6068redact