Pseudonymisation or anonymisation: what is the difference?
Anyone who shares documents containing personal data soon comes across two words: anonymisation and pseudonymisation. They are often used as if they mean the same thing, but under the GDPR the difference is significant. Anonymous data falls outside the GDPR; pseudonymised data is still personal data. This article explains what both terms mean, where they appear in the law and when to choose which.
What is pseudonymisation?
The GDPR defines it in Article 4(5). In short: you process personal data in such a way that it can no longer be attributed to a specific person without the use of additional information. That additional information (the "key") must be kept separately and protected by technical and organisational measures.
An example: in a letter you replace "John Smith" with [PERSON_1] and "GB33 BUKB 2020 1555 5555 55" with [IBAN_1]. Someone who only sees the letter does not know who it is about. But you keep a list stating that [PERSON_1] is John Smith. With that list you can always restore the data. That is pseudonymisation.
What is anonymisation?
"Anonymisation" is not defined in the articles of the GDPR, but it is addressed in Recital 26. It states that the GDPR does not apply to anonymous information: data that does not (or no longer) relate to an identifiable person. Whether someone is identifiable is assessed by taking into account "all the means reasonably likely to be used", such as cost, time and available technology.
True anonymisation is therefore irreversible: there is no key, and the person cannot be identified by combining data either. Europe's data protection authorities (then the Article 29 Working Party, in Opinion 05/2014) name three risks you must rule out:
- Singling out: can you isolate one individual in the data?
- Linkability: can you link records about the same person to each other or to another source?
- Inference: can you deduce something about a person, such as their job or where they live?
The difference at a glance
- Reversible? Pseudonymisation: yes, with the key. Anonymisation: no.
- Covered by the GDPR? Pseudonymisation: yes, it remains personal data (Recital 26). Anonymisation: no.
- Used for? Pseudonymisation: when you will need the real data again later. Anonymisation: publication, statistics or examples.
Why pseudonymisation is still valuable
Because pseudonymised data remains subject to the GDPR, some people think it is pointless. It is not. The GDPR explicitly mentions pseudonymisation as a good safeguard: in Article 25 (data protection by design), Article 32 (security of processing) and Article 89 (safeguards for research and statistics). It limits the damage if a document ends up in the wrong place: without the key, the recipient sees no names, numbers or addresses.
How VerumVault does it. VerumVault replaces names, ID numbers, IBANs, addresses and phone numbers with labels such as [PERSON_1]. The link between label and real value (the "Token vault") exists only in your own browser. So you share a pseudonymised text and keep the key yourself. After the AI has answered, you restore the real data locally.
Careful: labels alone are not always enough
A text can give someone away even without a name. "The only female councillor in a small village" or "our director who had a heart attack in March" is instantly recognisable to insiders. Always reread a pseudonymised text for:
- job titles, roles and unique characteristics;
- dates and events that point to one person;
- small towns, street names and company names;
- health, religion or other special category data (Article 9 GDPR).
Replace such details with something more general ("a councillor", "in spring") or remove them if the recipient does not need them. That reflects the principle of data minimisation in Article 5(1)(c) GDPR.
Which method should you choose?
- Having a letter checked by AI or an adviser: pseudonymise. You want to use the answer with the real names afterwards.
- Publishing a judgment, case study or example: anonymise. Delete the key and remove indirect identifiers too.
- Internal analysis or research: often pseudonymise, with the key held by a different person or department.
For practical steps, read Anonymise a document in 3 steps and How to anonymise a letter under the GDPR. Using AI tools? Also read How to use ChatGPT in a GDPR-compliant way.
Anonymise your document for free with VerumVault →Runs locally in your browser · no upload · no accountFrequently asked questions
Is a pseudonymised document still personal data?
Yes. As long as someone can link the data back to a person using a key or other information, it remains personal data under the GDPR (Recital 26).
Does an anonymised document fall outside the GDPR?
Yes, if the anonymisation is truly irreversible and the person cannot be identified by any other means reasonably likely to be used. Then it is no longer personal data.
Is replacing names with labels anonymisation?
Usually not. If you keep the link between label and name, it is pseudonymisation. Only when the key is gone and the content cannot identify anyone either is it anonymisation.
Where is pseudonymisation mentioned in the GDPR?
The definition is in Article 4(5). Pseudonymisation is named as a safeguard in, among others, Articles 25, 32 and 89 GDPR.