# Anonymisation

> Removing or masking the details that identify people, so data can be used or shared without exposing anyone.

**Category:** Safety

**Anonymisation strips the identifying details out of data.** Names, emails, and ID numbers come out (or get masked), leaving the useful shape behind, the trends, counts, and patterns, without the people. Done well, no one can be traced from what remains.

It is the safe move before sharing data with an AI tool when you only need the analysis, not the individuals. Note the weaker cousin: **pseudonymisation** swaps identifiers for codes but can be reversed, so it is not the same as true anonymisation.

## At work

Want Claude to spot trends in support tickets? Replace customer names and emails with 'Customer 1, 2, 3' first. You keep the insight, lose the risk.

## Often confused with

Pseudonymisation. Anonymisation cannot be reversed; pseudonymisation swaps in codes that can be traced back, so it is less protective.

## Related terms

- Personal data
- GDPR
- Enterprise tier

## See it in a lesson

[AI at work: safety and privacy](/en/learn/foundations/ai-at-work-safety-and-privacy)

## Further reading

- [Data anonymization (Wikipedia)](https://en.wikipedia.org/wiki/Data_anonymization): What anonymisation is and how it is done.
- [Pseudonymization (Wikipedia)](https://en.wikipedia.org/wiki/Pseudonymization): The reversible cousin, and why it is weaker.
- [Data protection explained (European Commission)](https://commission.europa.eu/law/law-topic/data-protection/data-protection-explained_en): How this fits EU data-protection rules.

---

Last updated: 24/07/2026  
Source: Rollo Academy glossary