The change in mentality for data and network security solutions has made the shift from focusing on creating an impenetrable barrier to monitoring the activity under your control and investigating aberrations. The value of this strategy and its implementation was demonstrated earlier this year by MasterCard’s Safety Net.
In the three attacks this year, directed against two U.S. banks and one bank in South America, Safety Net identified anomalies such as large cash withdrawals or transactions outside the usual geographic area for a given account. MasterCard notified the banks and rejected transactions, limiting losses to less than $100,000 in each case, said Ajay Bhalla, president of enterprise security solutions at MasterCard. The company declined to identify the banks, citing confidentiality agreements.
Mercator Advisory Group believes the above example exemplifies the increased level of communication and collaboration that is developing across the payment ecosystem to more effectively curtail the activities of “bad” players.
Machine learning helps Safety Net discern a reasonable change in customer behavior from an aberration that signals trouble, Mr. Bhalla said. Layers of algorithms work together. For example, one spots a transaction unusual for a specific account. The next analysis delves into what else has been going on with the account and another may compare that activity with customer activity locally and regionally. A MasterCard-wide analysis can pick out similar unusual activity worldwide. “If there are several accounts doing same thing, then escalation happens and we start calling banks. This indicates more organized effort.”
Further, as machine learning becomes more widely used, the identification of patterns and the effectiveness of predictive analytics will come into play and help to further ensure the authenticity of data, and generate a framework for the next generation of trusted transactional engagement.
Overview by Joe Walent, Senior Analyst, Emerging Technologies Advisory Service at Mercator Advisory Service
Read the full story here