This refers to Bank Watch titled “Of card cloning” by Sajjad Bazaz. This is a very outdated problem for which solutions have existed since 2001. Card fraud is rising at an alarming rate. Fraudsters, typically highly sophisticated international organized crime rings nowadays, are using increasingly advanced methods to obtain and use consumer information to commit payment fraud. Alongside conventional credit and debit cards, we are seeing new pervasive payment channels such as prepaid cards, e-payments and now m-payments. These enhanced payments channels generate fresh opportunity for card issuers and merchants to infiltrate new consumer markets, however, in doing so, they are also exposing themselves to new fraud risks. The highly informed fraudster will be keen to shift his efforts to rapidly exploit emerging and vulnerable payment channels and less sophisticated payments markets, meaning patterns can be expected to change more quickly in the future and therefore the need for a dynamic fraud detection system will grow. Fraudsters will alter their behavior to avoid establishing fraud patterns detectable by automated card transaction analysis systems. This highlights the importance of having a readily adaptable fraud detection solution which can incorporate evolving fraud patterns and customer profiling strategies to identify transaction anomalies. Visa and MasterCard have already accelerated the migration from magstripe to chip cards and chip technologies in order to address concerns about escalating fraud levels. Under the scheme, the investment has been allocated to support a range of initiatives targeting Visa card issuing banks, acquiring banks, retailers and technology providers.

As a result of the liability shift, retailers who deploy chip reading technology stand to benefit from a shift in transaction liability through their acquirers. Card issuers will become responsible for fraudulent transactions if they have not implemented chip, where it can be demonstrated that chip would have prevented the fraudulent transaction from occurring. Retailers are also expected to benefit from the ability to utilize remote payment technologies and to add proprietary loyalty schemes through partnerships with the banking community. Many Adaptive Classification Engine based solutions provide intelligent fraud detection, generating fraud alerts by executing mathematical models against incoming transactions to compute a fraud score. Uniquely, the ACE intelligent detection model is able dynamically self optimize and so detect emerging fraud patterns between model retrains, enabling a much higher detection rate to be maintained than is possible with, say, neural network methods, the detection performance of which tends to drop off shortly after a new model is deployed.

(Feedback at taseer@rangrez.info)

Lastupdate on : Thu, 15 Jul 2010 21:30:00 Mecca time
Lastupdate on : Thu, 15 Jul 2010 18:30:00 GMT
Lastupdate on : Fri, 16 Jul 2010 00:00:00 IST