Scrub the obvious personal data out of text before it goes somewhere it shouldn't: logs headed to a bug report, tickets pasted into a chat, prompts sent to an LLM. Seven detectors cover emails, phone numbers, US SSNs, credit cards (Luhn-checked, which removes most false positives without eliminating them), IPv4/IPv6, and IBANs — with per-category toggles and your choice of labeled or length-preserving replacement.
Everything runs in your browser; nothing is uploaded. For the metadata side of the same problem, the EXIF remover strips location data from photos and the PDF redactor handles documents; the fake data generator is handy when you need realistic-but-safe replacements.
Frequently asked questions
What counts as PII?
Personally identifiable information is any data that can identify a specific person — directly (name, email, SSN, phone number) or in combination (IP address plus timestamp, card number plus expiry). This tool targets the well-structured kinds that follow recognizable patterns: emails, phone numbers, US SSNs, credit card numbers, IPv4/IPv6 addresses, and IBANs.
Is my pasted text uploaded anywhere?
No. Detection and redaction run entirely in your browser with regular expressions and checksum math — the page makes no network request with your text, which is exactly why it is safe to paste logs or tickets that contain sensitive values.
How accurate is the detection?
Good for well-formed patterns, imperfect by nature. Card numbers are checksum-validated (Luhn) and IBANs mod-97-validated, which removes most false positives but cannot remove all of them: about one in ten random 16-digit numbers passes Luhn by chance, so a long numeric id — an order number, a transaction reference — can still be flagged, and a spaced numeric id can even be partially matched. SSNs must be dashed and pass format rules, and phone matching requires separators so plain digit runs are not flagged. Free-form PII — names, street addresses, dates of birth — has no reliable pattern and is not detected at all. Always review the output before sharing.
Can I use this to be GDPR or HIPAA compliant?
No tool can promise that, and this one does not try. It is a fast first pass for scrubbing logs, support tickets, and LLM prompts before sharing. Compliance regimes require process — data mapping, retention, access control — not just pattern redaction. Treat the output as a draft to verify.
What is the difference between the [LABEL] and mask modes?
Label mode replaces each finding with its category, like [EMAIL] or [CARD] — best when the reader should know what kind of data was removed. Mask mode replaces every character with an asterisk, preserving the original length — best when layout or alignment matters, such as fixed-width logs.