Recognises a check weaker than its rule
A check that lets through rows the rule forbids: one direction of a two-way rule, presence where the rule says exactly one, a range with an open end the rule closes, a value set wider than the rule's.
Can AI do this?
People still lead Few or none of the tested AI models showed they master this. 0 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 73% | Not shown |
| gemini-3.1-pro-preview* | Flagship | 44% | Not shown |
| gpt-6-astra* | Flagship | 100% | Mastered |
| claude-fable-5-1* | Flagship | 100% | Mastered |
| gemini-3.8-flash* | Mid tier | 76% | Not shown |
| gpt-6.1-sol* | Mid tier | 99% | Mastered |
| claude-sonnet-5-5* | Mid tier | 99% | Mastered |
| gemini-3.5-flash-lite | Low cost | 69% | Not shown |
| gpt-6-luna | Low cost | 74% | Not shown |
| claude-haiku-5-5 | Low cost | 74% | Not shown |
* From the earlier benchmark of 2026-10: 10 data-modelling test tasks, 3 runs per task, graded by claude-opus-5-5. These models have not taken the current 123 tasks yet; the two methods differ, so compare their figures with care. They are left out of the “most models” counts and the best-model picks.
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 check, the rows it lets through and the missing check
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: Rule: each account has exactly one IBAN and no IBAN is shared. Check: account_identifier IS NOT NULL only. Duplicates pass Acceptable: Two checks: completeness on account_identifier, and uniqueness over it. Or the rule split into two at stage 3, each with its check
Also known as: Check weaker than its rule (data quality rule)
Related knobits
Builds on
- K-567 Explains relationship cardinality and optionality
- Q-352 Explains traceability from data rule to check
All 30 knobits of Derive data quality rules.
Source: Review checklist 03-dq-rules (data-quality), row 11; Olson 2003; DMBOK2; formerly data-modelling K-711