Preventing Frauds with Fraud Detection System

Once online merchants start accepting online credit card payments, chargebacks and disputes are inevitable. As a payment gateway, Midtrans offers the 3DSecure as a form of protection, however 3DSecure does not guarantee a 100% fraud-free transactions. Banks and Principals (i.e. VISA, MasterCard, JCB, AMEX, etc) require merchants to maintain their own fraud ratio to be below Banks and Principals’s acceptable threshold.

Midtrans’ FDS are integrated directly into the payment flow, intercepting possible suspicious transactions as a second layer protection before the transactions are charged to the banks. The Fraud Detection System utilizes a combination of rule engine and machine learning algorithm. Our rule schemes are developed based on the prevailing fraud patterns in each industries, updated continuously based on the fraud trends. On areas where our rule engine falls short, our machine learning algorithm --a ‘smart’ machine which self-learns from historical transactional behaviors-- covers for it. Midtrans’s FDS is also built with a comprehensive blacklist database on top of our rule engine and machine learning algorithm, which we amassed from various sources (banks and principals’ fraud reports, known dispute cases, and our own monitoring activities) across the years. Every transactions that go through Midtrans are screened with all these mechanisms - which creates a robust fraud screening process.

To be able to catch fraud patterns without rejecting good transactions, Fraud Detection System relies on the data which merchants send to us - the more areas of data that we received, the more sophisticated the rule schemes which can be implemented. Hence, we encourage merchants to send us as much data as possible (your data would only be used for this purpose only - we take our merchant’s data security and privacy seriously).