From Isolated Incidents to Actionable Intelligence: Tackling Refund Abuse Early

Refund abuse and first-party misuse are outpacing existing controls — and too often sit below evidential thresholds needed for shared fraud databases owing to challenges in collecting robust evidence – allowing organised refund abuse to go largely unchecked. What if industry could share factual occurrences of refund claims and disputes that don’t meet formal fraud thresholds, and then derive cross-ecosystem risk signals from their velocity, linkage, and patterns? This session presents a merchant-first blueprint for sharing refund-abuse occurrences below fraud thresholds: the data model, risk assessment criteria, and governance that make sharing useful and compliant; how signals travel across retailers, issuers, and PSPs; and how merchants could use them to prevent losses without degrading customer experience. Alongside this, Lloyds Banking Group will share insights from their newly established dedicated team focused on refund abuse, including patterns they have observed and how these insights inform issuer-side strategies. We’ll draw on Cifas’s internal design options and discovery work with retailers and industry bodies, plus soon to be published University of Portsmouth research commissioned by Cifas into “refunder” behaviours. Attendees will leave with a practical framework to evaluate pilots in 2026: what to capture, how to share, and how to turn shared occurrences into decisions. No product pitch — just a clear path to an industry signal scheme that complements case-based sharing and improves outcomes for merchants. Key Takeaways: Understand why sharing refund-abuse occurrences below fraud thresholds surfaces patterns sooner. Adopt a merchant-ready taxonomy for refund-abuse scenarios tied to decision points. Apply privacy-by-design governance (risk assessment criteria, lawful basis) to keep sharing proportionate. Design an MVPpilot (minimal fields, KPIs, feedback loops) to validate lift. Translate academic insights on refunderbehaviour into policies and rules.

Some content is hidden, to be able to see it login here Login