Humans in the Loop: Why AI Can’t Replace Your Fraud Strategy

In this episode of Fraudology, host Karisse Hendrick is joined by a guest she has been trying to bring back to the show for years: Holly Sandberg, a veteran leader in fraud operations, Americas Advisory Board at MRC and the former Director of Trust and Safety at Reverb. As the industry grapples with the rapid rise of Generative AI, Holly and Karisse cut through the "robot takeover" anxiety to provide a grounded, expert perspective on the future of fraud leadership.

The conversation explores the evolving mechanics of AI in trust and safety, detailing why the "human element" remains the ultimate force multiplier for technical models. Holly shares why senior fraud leadership cannot be replaced by LLMs, noting that the critical "domain expertise" required to manage sophisticated fraud—and spot model drift—is not found in open-source data.

We also explore the "hot topics" dominating the fraud landscape today:

• The AI Visibility Strategy: Why volunteering for AI steering committees and working cross-functionally is the best way to prove your indispensability to leadership.
• Hallucinations and Data Gaps: A deep dive into the "bias toward certainty" in AI, featuring a hilarious yet cautionary example of "pizza fraud" hallucinations and the risks of depending on LLMs for math and complex policy.
• Career Resilience in a Changing Market: Tactical advice on navigating layoffs, the "quiet rehiring" trend, and how to pitch your value for a promotion by focusing on what you will do, rather than just what you’ve done.