Arabic PII Redactor
Paste Arabic or Arabic-English text, see the personal data that was found, choose how to replace it and copy the result.
Everything runs in your browser. Your text is not uploaded, stored or logged.
The Arabic text has not yet been reviewed by a native Arabic writer; the English governs.
Text
Examples (generated values, not real personal data):
Generated examples; not real personal data; any match to a real person is coincidental.
What was found
Redacted text
Details
| Type | Text | Score | Check | Country |
|---|
What it finds
| Type | How it is checked |
|---|
Person names are not detected on this page. The Python package finds Arabic names with an optional model.
How it works
- Arabic-Indic and Persian digits, right-to-left marks and Arabic separators are read through a copy of the text in which each character is replaced by exactly one character, so every span points into your original text.
- Candidates are checked: Luhn check digits for Saudi and Emirati ID numbers, the IBAN registry and MOD 97-10 for IBANs, birth dates and governorate codes for Egyptian numbers, and the published number plans for phone numbers.
- Weak formats (Kuwaiti and Bahraini numbers, plates, phone numbers written without a prefix) count only next to a context word such as جوال, هوية, لوحة, phone or plate, including forms such as وجوالي and بالهوية.
- The page runs a JavaScript port of the rule-based recognisers of the Python package; both give identical spans and scores on the 1,300 sentences of the evaluation set and on 3,000 random test texts.
- Detection is a technical aid: it does not prove that a text holds no personal data. Generated values pass the same checks but are random, and could coincide with a real number.