DOCX files contain more than visible paragraphs
Names and identifiers can appear in tables, headers, footers, hyperlinks, document properties and embedded images. A text-only pass can therefore leave private details behind.
AnonymizeDocs reconstructs the exported DOCX from the local package and preserves supported document structure. Every replacement still needs human review.
How to review a DOCX before AI upload
Select the DOCX, review the detected values and inspect the generated preview. Pro and Pack users can export DOCX; the Free plan is limited to PDF exports even though local OCR is available for supported files.
- Check paragraphs, tables, headers, footers and repeated placeholders.
- Review embedded PNG and JPEG images that may contain names or identifiers.
- If image OCR cannot be completed safely, export is blocked instead of silently preserving the original image.
- Open the exported DOCX in Word or another viewer and search for known source values.
Keep a source copy
An AI-ready copy is for analysis, not a replacement for the source record. Store and share the original according to your normal document-control process, and share only the minimum information needed for the AI task.
Important: AnonymizeDocs is a document preparation tool, not legal advice or a guarantee that every sensitive value has been found.
Continue exploring
Supported files
Review the current format matrix.
PII checklist
Identify private values before sharing.
See DOCX plans
Compare Pack and Pro export access.
Local document workspace