THANK YOU FOR SUBSCRIBING
"The solution learns from, and adapts to, suspected fraudulent transactions in an average of five milliseconds, instantly closing the gap for potential fraud, using the same card or device, across all members of the network"
Leading the payment industry with over 15 years of experience, CA Technologies co-created the 3-D Secure protocol in partnership with Visa to protect online e-commerce transactions when e-commerce was in its infancy. Since then, CA has been dedicated to perfecting the balance of robust security with customer experience to ensure that our customers can significantly reduce fraud losses, transaction abandonment and false positives. Furthermore, with a dedicated team of data scientists, CA strives to lead the online fraud detection market by leveraging machine learning techniques, artificial intelligence and neural network modeling. Our data scientists are experts with decades of experience in e-commerce, model building and big data.
Rich data network
CA Risk Analytics Network taps into the largest network of global cardholders and financial transaction data.
The CA Risk Analytics Network solution incorporates a new advanced neural network model, backed by unique real-time machine learning algorithms and networked device and entity reputation data, to protect 3-D Secure card-not-present transactions. It learns from, and adapts to, suspected fraudulent transactions in an average of five milliseconds, instantly closing the gap for potential fraud, using the same card or device, across all members of the network. The Network enables the Issuers to quickly assess the risk of a CNP transaction by analyzing data across multiple dimensions, including type of device, location, behavior and historical trends in the context of both the card and device behavior. Leveraging the neural network model, the solution offers zero-touch authentication, and card issuers can customize their approach to risk assessment by setting custom rules, establishing policies and designating events that require step-up authentication. The solution also employs a self-learning scoring model, which analyzes and compares historical and real-time transaction behavior. After which, Card and device behavioral profiles are immediately updated, influencing the model to reflect the most accurate risk score.
CA Technologies co-created the 3-D Secure protocol in partnership with Visa to protect online e-commerce transactions when e-commerce was in its infancy
By analyzing and comparing multiple dimensions of large-scale data—both recent and historical, across banks and geographies, the CA Risk Analytics network helps issuers detect anomalous behaviors for the cardholder, the device or both. Real-time learning and instant risk score update Card and device behavioral profiles are updated in real time to influence the network model so that it can accurately and quickly determine which transactions are risky and which aren’t.
Value for Customer
CA technologies was approached by a rather conservative bank as they wanted a little more return from their fraud-prevention strategy while ensuring their customer experience remained unaffected. First, CA technologies took a “before” snapshot of the bank’s strategy by measuring the bank’s fraud losses and TFPR (Transaction False Positive Ratio). The bank had 19.3 million U.S. dollars in attempted fraud in a quarter’s time. The bank was using a traditional neural network detection system to fight that fraud, reducing its losses to $10.9 million. For many banks, this level of success is acceptable as they can absorb that amount of loss. However, this particular client wanted to see if they could implement a more effective system to further reduce fraud losses without impacting the TFPR. Importantly, the CA Risk Analytics Network also afforded the client’s more choice in how they handled risk and the experience of their customers. By choosing to use the network to reduce the number of false positives from 178,797 to 116,218, the bank was able to save 62,579 genuine customers the annoyance and frustration of a falsely declined legitimate transaction. Today, like many institutions, this bank continues to explore a healthy mixture of both a reduction in fraud losses and false positives. With an approach like this one that’s coming at the problem in real time, modern issuers can expect to see an average reduction in fraud by 25 percent. Along with this dramatic improvement, other benefits include reduction in false positives, which can decrease by 35 percent. The company will also allow issuers choose a combination of both if required.
In the future, better and faster solutions are bound to be introduced into the market with the continuation of the technological arms race between data scientists and hackers as these both camps will keep pushing their boundaries of what’s possible. A significant part of this development will be driven by the incredible amount of data that will be produced by initiatives like 3-D Secure 2.0 and PSD2—it’s all a treasure trove of information that data scientists can further use to keep ramping up the sophistication of fraud detection and prevention techniques in the coming few years.