Securing ecommerce payments and fighting fraud with AI

This editorial was first published in our Web Fraud Prevention and Online Authentication Market Guide 2017/2018. The Guide is a complete overview of the fraud management, digital identity verification and authentication ecosystem provided by thought leaders in the industry from leading solution providers (both established and new players) to associations and experts.

In fact, the total number of ecommerce breaches, according to Experian, has increased 56% compared to 2016. Nevertheless, Artificial Intelligence (AI) and machine learning promise to radically shift the balance of power between merchants and the criminals who seek to steal from them.

The growth in digital payments and transactions has left merchants, PSPs and ecommerce companies vulnerable to sophisticated new cyberattacks. Furthermore, the number of people adopting the use of apps and mobile connectivity for making transactions is growing at an astounding rate. This is why, in part, technologies such as Artificial Intelligence (AI) and machine learning are critical in helping organisations fight fraud in better and more effective ways than ever before.

AI is an important development for the payments and transactions industry because merchant business models are evolving daily, from next day delivery of goods to digital downloads. Machine learning used in order to fight fraud is the logical solution for navigating this ever-changing landscape. Other anti-fraud systems using analytics that do not utilise machine learning capabilities, flag credit card purchases that take place outside a customer’s country of residence, for example, or unusual payments to remote suppliers.

The problem with such systems is that those rules are created by humans, who cumulate and combine data and intuition. It has been proven to be somewhat effective, although the approach can be costly, slow, leading to false positives, and failing to keep pace with emerging trends.

Machine learning detects fraud in real-time and learns from trends, identifying quickly emerging fraud patterns. By integrating and analysing changing, unstructured, and fast-moving data in ways that humans alone cannot do it, machine learning employs the running of multiple self-learning algorithms in parallel to increase fraud detection and prevent it. Additionally, it can identify rare or never-before-experienced fraud events, automate tedious tasks and provide an anti-fraud solution that allows merchants, PSPs and their customers to rest easy knowing they are being protected by a sophisticated approach.

Even though this is an important development in fighting fraud and it is true that machines can better perform the arduous task of evenly sifting through massive sets of structured and unstructured data for fraud patterns, it is still critical to note the role humans play and how company culture must support it.

This is even more so the case since commerce operates in an omnichannel environment across multiple devices and touchpoints. Bad experiences, such as chargebacks, caused by fraudulent activity increase and subsequent losses in online marketplaces, impact those touchpoints that connect buyers and sellers.

Cyber-criminals have familiarised themselves with the ins and outs of payment processes. According to the Nielsen report, fraudsters steal about 5.65 cents per every USD 100 spent. Occurrences of identity theft, phishing and account takeover are ever increasing. It stands to reason why credit cards are the most popular target for fraud. It does not take much to conduct a ‘card not present’ transaction for online payments. Moreover, the dark web has provided a platform for a relatively easy exchange of stolen data.

These hi-tech hackers have become savvy in detecting vulnerabilities in systems, and pinpoint those backdoors in order to compromise the system and commit fraud. They utilise distributed networks, big data and the dark web to locate these vulnerabilities and optimise their financial gains. In fact, they are devising multidimensional tactics that inflict damage by sequentially compromising more than one point of vulnerability.

Machine learning provides a powerful solution that is responsive and dynamic, user-friendly and easy to maintain. Legacy-based rules of anti-fraud systems are breaking down at this level of sophistication, speed and scale. They lack in performing analytics and delivering risk scores very efficiently. In addition, they are not typically operating in real-time and with the same level of accuracy.

Machine learning can help by acknowledging behaviour to achieve better and more effective decision-making. This can be applied when conducting identity verification, payment authorisation, checkout scoring and merchant underwriting. The underlining result is that it significantly reduces fraud loss and chargebacks.

Overall, online fraud is expected to continually evolve to keep pace with the rapid development of technology. All the constituents in ecommerce, from merchants and PSPs to financial institutions, must stay ahead of the curve in order to protect themselves. The ramifications of failing to do so can be grim, making it vital to realise and embrace the power of machine learning and AI technologies to detect and prevent fraud in all ecommerce channels.

About Fiona Brown:

Fiona Brown has operated within the field of Risk Management for more than 20 years. She has always worked within Payments and has experience both within Acquiring and PSPs, having held senior roles at both First Data Merchant Services and Pay360 (formerly PayPoint Online). She joined Credorax in April 2017 where she assumed responsibility for the Fraud and Risk teams.

About Credorax:

In 2007, Credorax saw an opportunity to change the landscape of traditional merchant acquiring by using its technology assets to address the needs of online merchants of all sizes. This led the company to evolve into a Merchant Acquiring Bank specializing in cross-border ecommerce, with more than 200 employees globally and operations spread across Europe, the US, UK, Malta, and Israel.