Assess whether explainability artifacts would survive an exam (5bb0d6)

Situation

Explainability artifacts would survive sits with model-risk officer because a new use case bolted onto a model approved for a narrower purpose hit an insurer scoring claims with a third-party model. Evidence is hiring-tool adverse-impact tables; write the AI Governance Inventory and Regulatory Fit option that extract can carry.

Decision

Model-risk officer in an insurer scoring claims with a third-party model must choose Policy or governance breach / Model defect / Dual failure / Hold for the missing fact using hiring-tool adverse-impact tables after a new use case bolted onto a model approved for a narrower purpose.

Hypotheses to test

• Hiring-tool adverse-impact tables reads as Policy or governance breach once a new use case bolted onto a model approved for a narrower purpose is lined up to the same AI Governance population.
• Hiring-tool adverse-impact tables is closer to Model defect after a new use case bolted onto a model approved for a narrower purpose; Policy or governance breach would over-claim this Inventory and Regulatory Fit extract.
• Dual failure is still live in hiring-tool adverse-impact tables for model-risk officer in an insurer scoring claims with a third-party model.
• Hiring-tool adverse-impact tables is missing the fact model-risk officer needs after a new use case bolted onto a model approved for a narrower purpose; stop this AI Governance close.

Analysis required

• Split policy-or-governance failure from a model defect using prompts, outputs, and human edits in hiring-tool adverse-impact tables.
• Reproduce the incident row in hiring-tool adverse-impact tables and say whether it ever touched production data.
• Split policy-or-governance failure from a model defect using prompts, outputs, and human edits in hiring-tool adverse-impact tables.
• For this AI Governance Inventory and Regulatory Fit file, read hiring-tool adverse-impact tables against a new use case bolted onto a model approved for a narrower purpose and write the one fact that would move explainability artifacts would survive for model-risk officer.

Recommendation

On this AI Governance / Inventory and Regulatory Fit decision, bind Policy or governance breach to hiring-tool adverse-impact tables after a new use case bolted onto a model approved for a narrower purpose: if hiring-tool adverse-impact tables supports it, model-risk officer executes Policy or governance breach in an insurer scoring claims with a third-party model and logs the owner. If hiring-tool adverse-impact tables instead fits Model defect, take that path. Where hiring-tool adverse-impact tables after a new use case bolted onto a model approved for a narrower purpose is silent, model-risk officer must leave the classification unresolved and name the missing control or provenance fact.

Command returns

• Bottom-line AI Governance option on explainability artifacts would survive, then the evidence in hiring-tool adverse-impact tables, then the action for model-risk officer
• Hypothesis scorecard against hiring-tool adverse-impact tables: supported / rejected / untestable
• Missing page in hiring-tool adverse-impact tables after a new use case bolted onto a model approved for a narrower purpose, if any
• Regulatory or exam hook Inventory and Regulatory Fit would cite