The Office of the Comptroller of the Currency (OCC) Supervisory Guidance on Model Risk Management released in 2011 has set principles in practice however the challenges and technology continues to evolve.  

Financial Crime / Anti-money laundering (AML) efforts are all the more challenging to prevent crimes with sources stating significant increase in the frequency and creativity of the bad actors.  

• FinCen's analysis revealing ransomware "substantial increases" (FinCen 22)
• SEC's 9% yoy enforcement action increase (SEC 22)
• PYMNTS reporting 62% large banks report financial crime increases (PYMNTS 22-23)
• ASFCS' reporting significant increases in entity-centric financial crime strategy (ASFCS '22)
• INTERPOL cites Financial crimes as top global police concern (INTERPOL '22)
• Feature Space's ~60% fraud rate increase, however, FIs using AI/ML had lowest levels of financial crime, including fraud (State of Fraud/Financial Crime '22)

Financial Institutions (FI), in response, continue efforts to broaden and scale the efforts, technology and talent to develop, implement and use models while institutionalizing governance, policies, and controls.  Model Validation is as important to follow-through on the efforts and have a great opportunity to scale technology with AI/ML tools.  

A Model Validation Framework with applied AI/ML methodologies can enable the more rapid evaluation of  soundness and provide real-time evidence to build on and re-deploy.  AI/ML is built for ongoing monitoring, process heat-mapping of risks and ideal to measure benchmarks against historically quantified controls. Outcome, back testing, scenario concept use cases as well as testing, again, can be even further optimized with AI/ML.  

As Feature Space' State of Fraud/Financial Crimes '22 report states, "FIs using AI/ML had the lowest levels of financial crime"; AI/ML is a supportive best practice government agencies such as OCC reference.