Hawk, a financial crime detection technology firm, and AML Intelligence have published a new eBook drawing on the experiences of senior anti-financial crime practitioners to address growing challenges around AI model governance.
The publication, titled ‘Inside AI Model Management: Lessons from Anti-Financial Crime Leaders’, features in-depth interviews with five practitioners from ING, Rabobank, Apple Bank, and Credit Suisse.
It comes as AI model governance has emerged as a pressing concern for anti-financial crime (AFC) teams, driven by rising adoption of machine learning, generative AI, and agentic AI in anti-money laundering and fraud prevention efforts.
Despite widespread institutional enthusiasm for AI, with 91% of financial institutions now encouraging its use for financial crime and compliance, adoption has not been without friction. Research from Hawk and Chartis found that 83% of AFC professionals have experienced difficulty interpreting or trusting AI model outputs, while 70% have encountered model performance issues.
The eBook covers a range of governance topics, including why model development fails, the importance of problem definition and data quality, how AI and rule-based models can complement one another, and how institutions monitor for both model drift and human drift.
It also addresses the role of explainability, the “human in the loop” principle, and the distinct governance challenges posed by machine learning compared with generative and agentic AI.
Hawk, which helps financial institutions tackle financial crime through AI-driven solutions, argues in the report that the institutions best placed to lead on AI are not necessarily those with the most sophisticated models. Rather, it is those that construct governance frameworks which are resilient, auditable, and capable of adapting to emerging risks more quickly.
Copyright © 2026 FinTech Global
Copyright © 2018 RegTech Analyst
