Recognises an uncovered sensitivity or access requirement
Personal, confidential or regulated data with no access requirement, so classification, masking and row-level access are designed later from guesswork.
Can AI do this?
People still lead Few or none of the tested AI models showed they master this. 1 of 10 models master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 93% | Mastered |
| gemini-3.1-pro-preview | Flagship | 42% | Not shown |
| gpt-6-astra | Flagship | 84% | Not shown |
| claude-fable-5-1 | Flagship | 86% | Not shown |
| gemini-3.8-flash | Mid tier | 44% | Not shown |
| gpt-6.1-sol | Mid tier | 83% | Not shown |
| claude-sonnet-5-5 | Mid tier | 84% | Not shown |
| gemini-3.5-flash-lite | Low cost | 45% | Not shown |
| gpt-6-luna | Low cost | 45% | Not shown |
| claude-haiku-5-5 | Low cost | 44% | 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 data and the access requirement it needs
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: Customer personal identification codes mentioned in a data-item requirement and nowhere else Acceptable: "A customer's personal identification code is visible only to staff servicing that customer's accounts." type: access, sourced to the data protection policy.
Also known as: Sensitivity and access not covered (data requirement)
Related knobits
Builds on
All 47 knobits of Capture data requirements.
Source: Review checklist 01-requirements, row 29; Stage 1 (sensitivity and access)