Why Banks Are Rethinking Model Validation with AI

Video

With AI and machine learning reshaping risk management, banks must enhance model validation frameworks. Rita Gnutti from Intesa Sanpaolo shares insights on integrating AI-driven validation, mitigating risks, and ensuring regulatory approval while upskilling internal teams.

Mar 21, 2025

Rita Gnutti, Executive Director Internal Validation and Controls, Group Chief Risk Officer Area, Banca Intesa Sanpaolo

Tags: Model risk AI and Technology (including Fintech)

The views and opinions expressed in this content are those of the thought leader as an individual and are not attributed to CeFPro or any other organization

Banks are increasingly integrating AI and machine learning into their risk management frameworks, but this shift brings both opportunities and challenges. To enhance model risk management, financial institutions are developing specialized assessment frameworks that account for the complexities of AI-driven models.

However, the adoption of these techniques requires significant upskilling of internal teams, ensuring that expertise remains within the organization rather than relying solely on external support. Mitigating risks associated with AI is a priority, with banks implementing safeguards such as closed knowledge bases and mandatory human oversight to maintain control over model outputs.

As AI-driven models are used for both managerial and regulatory purposes, institutions must navigate strict validation processes, balancing innovation with compliance to gain approval from regulators and top management.

Rita Gnutti Bio

Rita is “Head of Internal Validation and Controls” in Group CRO Area of Intesa Sanpaolo with responsibility on Internal Validation, Model Risk Management and II level credit controls and data governance controls at group level; she has been 15 years in Financial Risk Management as Head of Market and Counterparty Risk Internal Models. Main achievements in current role have been: Creation of a Group Model Risk Management Unit, Definition of MRM Group Policy, Model Risk Assessment Methodology, Model Tiering Methodology, Model Risk Appetite, Top Management Model Risk Reporting, Independent Validation for regulatory and managerial models, Evolution of Internal Validation on regular assessment of Data Governance Framework according to ECB guide on RDARR.