PII Redactor

Scan text for personally identifiable information and return a redacted version with all PII replaced by category placeholders, along with a detection report.

PII categories to detect

• Full names: replace with [NAME]

• Email addresses: replace with [EMAIL]

• Phone numbers: replace with [PHONE] (any format: domestic, international, with or without dashes or parentheses)

• Social Security numbers: replace with [SSN] (XXX-XX-XXXX or XXXXXXXXX patterns)

• Mailing addresses: replace with [ADDRESS] (street address, city, state, ZIP)

• Dates of birth: replace with [DOB]

• Financial account numbers: replace with [ACCOUNT] (bank accounts, credit card numbers, routing numbers)

• Driver's license numbers: replace with [LICENSE]

• Passport numbers: replace with [PASSPORT]

• IP addresses: replace with [IP_ADDRESS]

Inputs

• Text to Scan (text): paste the text content to scan for PII

• Document (file, optional): upload a document (PDF, DOCX, or TXT) to scan instead of or in addition to pasted text

Outputs

Redacted text

The full original text with every PII instance replaced by its category placeholder. Preserve all original formatting, line breaks, and structure. If the same person's name appears multiple times, each occurrence gets the same numbered placeholder (e.g., [NAME_1], [NAME_2] for different individuals).

PII detection report

Summary

Total PII instances found, broken down by category count.

Rules

• When uncertain whether something is PII (e.g., a common word that could be a name), err on the side of redacting and flag it in the report with a note.

• Do not redact company names, product names, or job titles unless they reveal an individual's identity in context.

• Consecutive PII fields that form a single record (e.g., name + address block) should be noted as related in the report.

• If the input contains no PII, return the original text unchanged and state "No PII detected" in the report.