The Current State of AI Governance and Model Risk Management

In the fast-paced domain of artificial intelligence (AI), governance, oversight, and model risk management are center stage and at the top of minds for business executives. The present state of AI governance and risk management is undergoing development, but it is also one of inconsistency and growing pains. As organizations continue integrating AI systems into their operations to automate decisions and business processes, the need for responsible oversight has never been more important. In fact, organizations across industries are now starting to grapple with the complexities of establishing AI governance frameworks. Many face the challenges of pulling a community of stakeholders together to implement even the foundational components of an AI governance model that aligns with their AI operations. AI governance frameworks are the go-to source and strategy to ensure their AI systems are developed and deployed ethically, transparently, and with a clear understanding of the potential risks. However, AI governance and model risk management are not new, especially within the financial services industries (FSIs). Many FSIs began to address AI governance and model risk management in the early 2010s, with substantial momentum within the last five years. In contrast, other industries have moved much slower over the same five years, taking a “crawl” approach as they come up to speed in adopting and instituting cross-functional teams dedicated to AI ethics and transparency. We are seeing more organizations employ governance-savvy data scientists, add internal and outside legal counsel, experts, and AI ethicists to their teams to navigate the landscape of explainability, bias prevention, risk mitigation, and positive societal impact.