GovernanceOS home: Defensible AI Platform Northline Advisory
NA

Signed in as Practitioner, Northline Advisory.

AccountHelp and supportSign out

Nebula MutualDefensible AI Library / E-23 Studio

E-23 twelve-question check

Prototype states

E-23 twelve-question check

Filled from Nebula Mutual’s records on 2026-10-16. Blanks say what to add. Any direct edit is marked typed.

typed stated in the records, as declaredcomputed from typed values counted here from typed recordsnot supplied or not in this tier

#QuestionFrom the recordsSource
01Is every system that meets E-23's definition of a model either on your model inventory with a risk rating, or exempted as negligible under a process that approves and tracks the exemption?6 systems on the inventory, a tier on 6 of 6; no exemption process recordedcomputed from typed values
02Does your model inventory carry the seventeen fields E-23 sets out, the version and the approver among them?12 of 17 fields supplied; missing version, reviewer, dependencies, approved uses, limitationscomputed from typed values
03Is each system's rating recorded with each factor and the fact that moved it, the tier, the date, the name of the person who set it, the date it is next re-assessed and the trigger events on which it is reviewed sooner?not supplied Tier and date are typed; add the factors, the next re-assessment date and the trigger events.not supplied
04Does the approval record for each system name one person, cite the validation report by an identifier that would retrieve it, state what the system is approved to do and until when or under what trigger, and carry the version of the policy that was in force?not in this tier Approval records with a named approver are kept in the institution’s own tenant. The file declares gate states only, such as “G4 passed 2026-08-30”, typed.not in this tier
05Does each validation report state, in the same document, what was not established: the properties the evidence did not cover, the changes it did not anticipate, and the parts of the system that were not in the room?not supplied Record a scope limitation on each validation report.not supplied
06For each system, is the review independent from development, and is the person accountable at G2 never the person accountable at G3?not supplied The file names no G2 validator and no G3 approver. Add both roles for each gated system.not supplied
07Does each vendor model's contract require notice before the model is retrained or rescaled?not supplied Add the contract record for each of the 3 vendor systems.not supplied
08Could your institution compute each vendor model's performance by segment from records it already holds?not supplied No segment-level performance records are held.not supplied
09Does every row of your trigger register carry a threshold, an owner and a consequence, all three fixed before the system was deployed?not supplied The file has no trigger register. Add a threshold, an owner and a consequence for each system.not supplied
10Does your signal register keep the signals closed without becoming incidents, each with the reason and the name of the person who closed it?not in this tier The institution’s own tenant holds the signal register.not in this tier
11For a decision one of your systems made on a given day, can the record show what it saw, what it drew on and which version produced it?not in this tier Decision-level records come from the institution’s own systems, through its tenant. None of the six can be shown from a workspace.not in this tier
12For each agent, can you show one recorded refusal of a verified claim, from production or from a seeded test?none recorded Seed a test that the assistant should refuse, and record the refusal.not supplied
Learn more about this check's source

Questions as published in the E-23 readiness check, version 1.0, 6 October 2026. Answers are drawn from this client’s records; blanks say what to add.