The PII redactor removes personal data from text before you paste it into ChatGPT, a support ticket or a bug report. It finds emails, phone numbers, card numbers, IBANs, US Social Security numbers, IP addresses, URLs and any names or words you add, highlights each one for review, and replaces them with labels, blocks, partial masks or consistent pseudonyms. Card and bank numbers are checksum-verified to avoid false positives. Your text never leaves the browser.
| Style | Card | |
|---|---|---|
| Labels | [EMAIL] | [CARD] |
| Blocks | ███████████████████ | ███████████████████ |
| Partial | j•••••@example.com | •••• •••• •••• 1111 |
| Pseudonym | [email protected] | [CARD_1] |
Emails, phone numbers, payment card numbers, IBANs, US Social Security numbers, IP addresses, URLs and any words you list, such as names.
Not usually. Card numbers must pass the Luhn checksum, which most random numbers fail.
Yes. Use pseudonyms — the same email always becomes the same [email protected] — or partial masks that keep the last four digits.
Only names and words you list. Free-text names and street addresses can't be detected reliably by patterns, so add them and review the result.
No. Detection and redaction run entirely in your browser.
The redactor runs a pattern for each kind of personal data over the text and resolves overlaps so each span is redacted once, in priority order: emails, URLs, IBANs, card numbers, US Social Security numbers, IP addresses, phone numbers and your own word list. Card-shaped numbers of 13 to 19 digits must pass the Luhn checksum and IBANs the ISO 13616 mod-97 check, so order numbers and other long digit runs are left alone. SSNs exclude the invalid 000, 666 and 9xx area numbers. Phone numbers must contain 7 to 15 digits. Custom words such as names or project codes are matched case-insensitively as whole words, longest first. Every match is highlighted in a review pane where clicking it keeps it unredacted. Replacements can be type labels like [EMAIL], solid blocks of the same length, partial masks that keep the last four digits or the email domain, or consistent pseudonyms where the same email always becomes the same [email protected], so the text stays readable for an AI model or a colleague. Everything stays in browser memory.