Prepare screenshots before sharing
Find and cover common sensitive values in screenshots or photographed documents before sending them to another person.
Choose document image
JPG, PNG or WebP
Automatic detection can miss sensitive information. Always review the final image before sharing.
About AI Sensitive Data Redactor
AI Sensitive Data Redactor uses browser OCR plus automatic sensitive-value detection to cover likely private information before an image is shared.
Common use cases
Redact working copies of forms
Create a derivative image with likely phone numbers, email addresses, IDs or card-like numbers covered for review or demonstration.
Reduce accidental disclosure
Use automatic detection as a first pass when a document contains several fields that should not remain visible.
How it works
1
The browser runs OCR to identify text-like regions in the selected image.
2
Detected text is checked for common sensitive-value patterns such as email addresses, phone numbers and identifier-like strings.
3
Matching regions are covered in the generated image, and you review the result before downloading it.
Privacy & file handling
Redaction analysis is designed to stay on the device
The supported image-redaction workflow runs OCR and pattern detection in the browser rather than requiring the document image to be uploaded to Zavilo for analysis.
Limitations and important notes
- Automatic pattern detection can miss sensitive information or mark ordinary text incorrectly.
- Visual information that is sensitive but does not match a recognized text pattern may not be detected.
- Always inspect the final pixels before sharing; automatic redaction is an assistive step, not a guarantee of complete privacy.
Automatic redaction vs manual masking
| Feature | Strength | Important limitation |
|---|---|---|
| Automatic sensitive-data detection | Fast first pass for common text patterns | Can miss unusual formats or non-text private content |
| Manual mosaic or blur | You choose exactly which visible regions to hide | Requires careful manual selection of every region |
How to use AI Sensitive Data Redactor
- 1
Choose a document image.
- 2
Run sensitive-data detection.
- 3
Review every redaction carefully.
- 4
Download the redacted PNG.
Tips for better results
- Automatic detection can miss or misclassify content, so always review the result.
- For highly sensitive documents, manually verify every visible field before sharing.
Frequently asked questions
Does automatic redaction guarantee all private data is removed?
No. It assists with common patterns, but the user must review the final image before sharing.