1. Client-Side Processing Only

CleanRedact operates entirely inside your local browser memory. When you select or drag a document into the app, it is loaded into an in-memory ArrayBuffer via the HTML5 FileReader API. At no point is any file, page, or extracted text sent across the network or uploaded to any cloud server.

2. Local OCR & Facial AI

Text pattern recognition (IBANs, emails, phone numbers, amounts) and face detection run locally using WebAssembly and client-side machine learning (Tesseract.js and TensorFlow.js BlazeFace). No document content, facial coordinates, or biometric information leaves your computer.

3. Destructive Flattening

Redactions are permanently burned into raw image pixels during the export process. The underlying vector text layer and all EXIF / document metadata (GPS coordinates, device serials, author history) are stripped completely so that no information can be recovered.

4. Zero Tracking & No Cookies

CleanRedact does not set tracking cookies, does not collect telemetry, and does not integrate third-party advertising or analytics networks.

5. GDPR / DSGVO & Compliance

Because zero personal data is transferred to or processed by our servers, CleanRedact naturally fulfills strict GDPR/DSGVO requirements (including Art. 25 Privacy by Design and Art. 32 Security of Processing) and is safe for use in confidential banking, medical, and legal environments.