Recognises production personal data in non-production
A test environment loaded with an unmasked production copy.
Can AI do this?
AI does this partly Some of the tested AI models master this; others do not show it. 2 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 91% | Mastered |
| gemini-3.5-flash-lite | Low cost | 34% | Not shown |
| gpt-6-luna | Low cost | 91% | Mastered |
| claude-haiku-5-5 | Low cost | 88% | Not shown |
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 defect named on the unmasked production copy outside production, and its fix or the question for its owner
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: EV-002 SIT, data_source: production, masking: [] Acceptable: Synthetic data, or masked production with an approved exception
Also known as: Production personal data in non-production
Related knobits
Builds on
All 15 knobits of Plan environments and capacity.
Source: Review checklist 02-environments-capacity, row 1 (27002 8.33; GDPR Art. 25)