Model risk management
Managing the risk of losses or harm from models that are wrong or misused.
It covers model inventory, development standards, independent validation, approval, ongoing monitoring and periodic review. Banking guidance such as SR 11-7 is a well-known example, and its ideas apply widely to AI.
On the exam
Independent validation, by people who didn’t build the model, is the core control.
Related terms
- Independent model validationValidation by people who didn’t build the model.
- AI inventoryA register of every AI system in use, with owner and risk rating.
- Model monitoringWatching a live model’s inputs, outputs and performance.
AAIA Test yourself on this with timed AAIA practice questions in the AAIA Exam Simulation Hub, or read the free AAIA exam guide.
Test yourself with today’s question of the day.
