Model Risk Meets AI: Bridging Gaps in Governance

As AI and machine learning (ML) become increasingly integrated into high-impact decision-making, the risk governance landscape is shifting. Traditional model risk management (MRM) frameworks, designed for conventional statistical models, now face new challenges presented by AI models – challenges such as complexity, limited transparency, and rapid change.

In this webinar recording, Crowe specialists explore how foundational MRM principles – governance, testing, and life cycle oversight – can be extended and adapted to suit the needs of modern AI systems. The session highlights key differences between traditional and AI and ML models and shows how existing risk frameworks must evolve to remain relevant and effective.

Crowe specialists examine how leading frameworks such as the National Institute of Standards and Technology AI Risk Management Framework (NIST AI RMF), ISO/IEC 42001, and state-level regulations like the Colorado AI Act are shaping expectations for AI governance. Viewers will gain a working understanding of how to navigate the intersection of compliance, transparency, and operational effectiveness across the AI model life cycle.

After watching this webinar recording, you should be able to:

• Describe how MRM principles support governance, validation, and accountability in AI model life cycles
• Distinguish between conventional and AI and ML models and evaluate how MRM must adapt to address unique risks such as complexity, data drift, and limited explainability
• Interpret emerging regulations and standards shaping AI risk oversight, including NIST AI RMF, ISO/IEC 42001, and the Colorado AI Act
• Apply a practical framework to manage AI model risk using strategies such as inventorying, risk tiering, documentation, and performance monitoring

Note: Only attendees of the live webinar are eligible for CPE or CLE credit for qualifying webinars. If you view the webinar recording on this page you may not be eligible for CPE or CLE. For questions about CPE, contact [email protected]. Additional CPE information.