After Processing Over 106 Million Disputes Last Year

Visa is introducing new AI tools to handle an increasing number of disputes, primarily arising from unrecognized charges on consumers’ increasingly complex statements. This issue has seen a roughly 35% increase over the past six years.

To address these challenges, Visa is launching six AI-driven tools. Three are aimed at helping issuers better analyze and centralize dispute data, while three others focus on merchants to improve data sharing between them and card issuers.

According to Don Apgar, Director of Merchant Payments at Javelin Strategy & Research, the traditional chargeback and dispute process is designed around consumers working with their card issuers and merchants with their acquirers. However, legacy data formats like ISO 8583 only allow for brief merchant descriptors without purchase details, making it difficult for consumers to remember transactions.

As a result, many disputes are initiated by consumers clicking the ‘dispute this charge’ button on mobile banking apps, hoping the card issuer can provide more information later. This process often leads to an overloaded legacy system where many chargebacks might be classified as friendly fraud, but in reality, many disputes occur due to timed-out inquiries.

Building a Process for Direct Communication

Because much of the dispute process remains manual, it struggles to scale with increasing credit card transactions and disputes. The solution involves creating a direct communication channel between card issuers and merchants.

Apgar suggests models such as shared databases where merchants upload information for issuer access or API hubs enabling issuers to query merchants and receive automated replies. Issuers can also use this data to proactively expand their cardholder statements, preempting consumer inquiries with detailed purchase information.