This handbook provides a practical governance framework for the adoption of Agentic AI in financial institutions, addressing the risks introduced by autonomy, adaptive behavior, and real-world execution. It demonstrates that Agentic AI does not require a new, AI-specific governance regime, but rather a more explicit, continuous, and disciplined application of Model Risk Management (MRM), fully integrated with Operational Risk Management (ORM) and embedded within the Three Lines of Defense (3LoD). By translating autonomy into tier-based controls, preventive, detective, and adaptive governance mechanisms, and operationally credible accountability, the handbook enables institutions to scale Agentic AI while preserving auditability, operational resilience, and supervisory confidence.

What you’ll get from this report

A clear, supervisory-credible blueprint to govern Agentic AI in practice — including tiering logic, integrated MRM and operational risk controls, and concrete banking and insurance use cases that translate governance principles into operational reality.
 

Read the full Practitioner Handbook to explore how financial institutions can scale Agentic AI without losing control. 

Sign up to access the full report.