What does anonymise mean? Definition, examples and techniques

To anonymise is to transform information that identifies a person so it can no longer be attributed to them — irreversibly, with no key to undo it. Once that is achieved, the data is no longer personal data and falls outside the GDPR (Recital 26).

Anonymised data: the exact meaning

Anonymised data no longer identifies a natural person, directly (name, ID number, phone) or indirectly (combining role, town, date, condition or amount with other sources). If re-identification is possible by reasonable means, the data is not anonymised.

Examples of anonymising a document

• "Mary Smith, ID 12345678Z, 4 Main Street" → "[NAME], [ID], [ADDRESS]".
• Council minutes where a resident speaks: keep the debate, redact their identifiers.
• An invoice used in a demo: remove tax ID, IBAN and address; keep amounts and line items.

Anonymisation techniques

1. True redaction of the text (not a black box drawn on top).
2. Masking with asterisks or [NAME]-style labels.
3. Generalisation: age 37 → "30-40"; postcode → city.
4. Aggregation: publish totals instead of individual records.
5. Synthetic data: replace with consistent fictional names and numbers.

Anonymising is not pseudonymising

Replace a name with a code and keep the mapping table and you have pseudonymised: still personal data, still under the GDPR.

Common mistakes

• Covering text with a black shape in the viewer — the text stays underneath.
• Forgetting PDF or Word metadata.
• Leaving indirect identifiers: case number, licence plate, scanned signature.
• Anonymising page one but not the annexes.

How to anonymise in practice

With anonimizia you upload the PDF or Word file, the system detects names, ID numbers, addresses, phones, IBANs and plates, you choose anonymisation (redaction, asterisks) or pseudonymisation (code, token, synthetic value) and download the public version with an auditable report.