A Chief Risk Officer is concerned that AI models are being deployed without adequate documentation of their limitations. Which policy control MOST directly addresses this concern?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — you need the model card or model documentation sheet because it forces known limitations, failure modes, and out-of-scope use cases to be documented before production. That gives risk, compliance, and operations a concrete artifact to review.
Full Explanation
Mandating a completed model card or model documentation sheet with known limitations, failure modes, and out-of-scope use cases before production deployment directly addresses the concern because it requires model constraints to be documented and communicated before use. It creates a structured artifact that risk, compliance, and operations teams can review. Requiring a minimum accuracy threshold measures performance but does not document limitations. Limiting the number of AI models per business unit is a portfolio management control, not a documentation control. Requiring AI ethics training builds developer capability but does not ensure specific documentation for each model.