Identity and names
People, passports, licences, tax identifiers, national IDs and privacy-context birth dates.
DETECTION TRANSPARENCY
Choose what anonymise.online protects.
The detector is not an all-or-nothing switch. You can enable or disable broad data categories before each anonymization, then review the result before export.
Contact and location
Email addresses, phone numbers, postal addresses, postcodes and location values when the surrounding text supports a privacy interpretation.
Financial
Payment cards, IBANs, bank routing/BIC information, labelled bank accounts and insurance policy identifiers.
Credentials and secrets
API keys, access tokens, passwords, private keys, signed URLs and other high-confidence credential patterns.
Technical identifiers
IP addresses, device identifiers, vehicle IDs, serial numbers and context-bound account identifiers.
Health identifiers
Context-bound medical-record and patient identifiers.
How the decision is made
anonymise.online combines deterministic format checks, contextual rules and offline named-entity recognition. Strong formats such as payment cards, API keys and IBAN-like identifiers use structural validation. Ambiguous language such as names and organizations is filtered more conservatively using surrounding wording, language and shape signals.
This is intentionally reviewable rather than presented as perfect classification. The review page shows what was detected and lets you add manual blocks before creating the restore map.
Practical default: leave all categories enabled unless you have a clear reason to keep one category visible. For example, you may choose to leave organization names visible while hiding people, contact information and credentials.