The Fraud Fighter's AI Playbook

Chapter 4. GenAI-Powered Deception Techniques

If you don’t know me by now,

you will never never never know me

Harold Melvin & the Blue Notes, “If You Don’t Know Me by Now”

Who Is This Chapter For?

This chapter is suitable as an intro for data and product folk of various levels of experience and background. It opens with a wide discussion about the role of fraudulent intent in defining deception, then digs into common methods for falsifying data across digital environments. The chapter can serve as background for planning your risk management strategy in the era of GenAI.

The chapter mostly draws examples from the realm of deception methods that impact digital media and trust (i.e., “How do I know that I’m looking at content/details that came from a genuine, trustworthy human being?”). If you’re looking for examples that focus on identity deception, you’ll find them in Chapter 5. Further examples of signal deception of all sorts can be found in equivalent training content for ecommerce and fintech, such as the “Measure What Matters” video series on FraudLab.com.

By the end of this chapter, you will be able to discuss key examples of deception techniques and understand how they are powered by GenAI.

Introduction

This chapter might just be the most pessimistic part of the book. Detection tactics have grown exponentially stronger with GenAI, to a point where every week, fraud fighters reveal new scheme variations. Fraud prevention has always been a game of Whac-a-Mole, but now ...