Recognises a RACI row contradicting the decision rights
The activity exercises a decision right and its A is not that right's decider. Two answers to "who decides".
Can AI do this?
Not settled yet The benchmark tested this, but the models' answers do not settle yet whether they master it. 0 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 74% | Not shown |
| gemini-3.5-flash-lite | Low cost | 72% | Not shown |
| gpt-6-luna | Low cost | 77% | Not shown |
| claude-haiku-5-5 | Low cost | 71% | 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 row, the decision type and the one answer to who decides
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: "Approve a change to loan term definitions", accountable: RO-009 CDO; DC-002 says the domain owner decides Acceptable: A = RO-003 Retail lending data owner; the CDO consulted
Also known as: RACI disagrees with the decision rights
Related knobits
Builds on
- G-253 Explains the single-accountable rule of a RACI
- G-252 Explains decision rights and the escalation ladder
All 26 knobits of Design the governance operating model and stewardship.
Source: Review checklist 02-operating-model, row 11 (DCAM; RAM)