The Fraud Fighter's AI Playbook

Chapter 7. Supercharging ML Models with GenAI

Some of them want to use you

Some of them want to get used by you

Some of them want to abuse you

Some of them want to be abused

Eurythmics, “Sweet Dreams Are Made of This”

Who Is This Chapter For?

This chapter is intended for fraud strategists, product managers, and technical teams working at the intersection of fraud prevention and AI deployment. It is especially relevant for those tasked with shaping policies, building detection systems, and guiding responsible adoption of GenAI tools within their organizations.

While it assumes familiarity with the operational realities of fraud detection—such as rules engines, anomaly detection, and case management workflows—it focuses less on technical implementation and more on strategic framing.

The chapter invites fraud fighters, both veterans and newcomers to the industry, to think about how the machine learning (ML) models of fraud prevention must evolve in a world where AI agents influence both sides of the cat-and-mouse game. Rather than presenting a technical recipe, it encourages a critical lens on automation, a careful weighing of trade-offs, and a reaffirmation of the human role in setting boundaries and values when designing detection ecosystems. This chapter focuses on applying this lens to ML models; the same analytical mindset will serve fraud fighters well in considering the possible application of GenAI more broadly in their work—a topic we cover in Chapter 8.

Introduction ...