Recognises an aggregate still identifying a person
Groups small enough to single out one customer, classified as non-personal.
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 | 96% | Mastered |
| gemini-3.5-flash-lite | Low cost | 40% | Not shown |
| gpt-6-luna | Low cost | 96% | Mastered |
| claude-haiku-5-5 | Low cost | 96% | 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: The aggregate and the group-size rule it needs
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: CO-011 branch_income_stats.avg_income grouped by branch and birth year, public; groups of 1 Acceptable: Minimum group size 10 with small cells suppressed, then internal
Also known as: Aggregate that still identifies a person
Related knobits
Builds on
All 16 knobits of Classify data.
Source: Review checklist 02-classification, row 3 (GDPR rec. 26; 20889)