Track each AI system through its governance lifecycle — from development through validation, production, monitoring, and retirement. SR 11-7 requires documented lifecycle management for all material models.
Model Validator or Model Risk Manager — tracking AI systems from development through retirement
WHEN
During model validation, before deployment approval, at scheduled review intervals, and at model retirement
WHY
SR 11-7 requires documented lifecycle governance for all models. Without tracked lifecycle states, the institution cannot demonstrate that models are under active oversight.
HOW
1. Review models pending validation 2. Approve or reject lifecycle transitions 3. Trigger re-validation when required 4. Archive retired models with audit trail