Are existing model risk management programs still "fit for purpose" in the age of AI?
​
While most Banks have evolved their model risk management programs to be consistent with the 2011 bank supervisory guidance (OCC 2011-12 and SR 11-7), there is still debate as to whether these programs remain “fit for purpose” given the rise of AI / ML methodologies.
 
In this primer, I directly address this "fit for purpose" question by performing a deep dive risk assessment of current AI / ML methodologies and development technologies, and overlaying the corresponding risks onto our existing model risk management framework.