Explains anonymisation against pseudonymisation
Hashed or tokenised identifiers remain personal data (GDPR recital 26); an aggregate is anonymous only if no group can single out a person (minimum group size, small-cell suppression; ISO/IEC 20889).
Can AI do this?
AI does this well Most of the tested AI models master this. 3 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 99% | Mastered |
| gemini-3.5-flash-lite | Low cost | 80% | Not shown |
| gpt-6-luna | Low cost | 99% | Mastered |
| claude-haiku-5-5 | Low cost | 99% | Mastered |
Mastered means a probability of at least 90% that the model has this knobit, estimated from its graded answers. “Not shown” means the answers do not settle it either way. How the models are tested.
What it looks like in practice
Expected result: Hashed or tokenised identifiers remain personal data (GDPR recital 26); an aggregate is anonymous only if no group can single out a person (minimum group size, small-cell suppression; ISO/IEC 20889).
How it is assessed: Answers a written question on the anonymisation and pseudonymisation correctly
Related knobits
Needed for
- S-203 Recognises an aggregate still identifying a person
- S-207 Recognises pseudonymised data treated as anonymous
All 16 knobits of Classify data.
Source: Practice the stage 2 reference and review checklist rest on