As AI and Real-Time Payments Converge, Anti-Fraud Efforts Must be at the Forefront

If fraud isn’t in focus, the benefits of the parallel rise of AI and real-time payments, and their intersection, have the potential to be rivalled by risks

By Damon Madden, Strategic Solution Consultant – Fraud, MEASA, ACI Worldwide

The Middle East is emerging as a global leader in the adoption and innovation of artificial intelligence (AI), a journey that began with the establishment of the world’s first AI ministry in the UAE. This region, characterised by rapid technological advancements, is poised to harness AI’s transformative potential, with economic predictions pointing to a significant GDP boost. According to McKinsey, AI could add an estimated US$150 billion to the GCC region’s economy, a figure that might soar even higher with the advent of generative AI. Concurrently, the Middle East is witnessing explosive growth in real-time payments, marking a 33.6% year-on-year increase in transaction volumes in 2023.

As these two powerful trends converge, the region stands at a crossroads, where the immense potential for economic and societal benefits is tempered by the critical need for robust anti-fraud measures to protect public trust. While this trend will no doubt have immense potential to unlock value for individuals and businesses, if not managed properly, it could also result in damage to public trust in both arenas. As regional transaction value hurtles towards the trillion-dollar mark (US$903 billion by 2028), the global financial ecosystem will certainly benefit. But just as real-time payments boost economies and enhance financial inclusion, they present opportunities for others who do not have our best interests at heart.

Challenges and opportunities

Legacy payment systems took days to process transactions, which allowed institutions more time to discover fraud. Real-time payments are great for the customer experience but also raise security challenges that are exacerbated by the speed and volume of transactions initiated, by fraudsters’ use of generative AI, and by the rush of enterprises to integrate real-time payment options as part of their digital transformation strategy. Interestingly, the very AI that intensifies the threat from fraudsters allows defenders more options to prevent illegitimate real-time payment transactions.

Saudi Arabia, having launched its sarie system only in April 2021, is already seeing a myriad of scams — investment, fake billing, remote access, identity theft — targeting customers of all ages. According to ACI Worldwide’s Scamscope 2023 report, the value of losses to Authorised Push Payment scams in Saudi Arabia is expected to increase by an annual compound rate of 26% between 2022 and 2027.

Middle East financial service industry (FSI) entities that integrate real-time payments must overcome a series of challenges. Take identity theft. Synthetic identity fraud is easier with AI, so in the spirit of “set a thief to catch a thief”, AI is the ideal tool to counter it. But to do so effectively requires an abundance of granular, high-quality data. The accuracy of AI models is determined by a range of factors, but most of them are data-related. With too few data points or too many inaccurate ones, the results will be too inadequate to stand as an effective countermeasure to fraud.

Fraudsters also need data to be effective, so the other side of the data challenge is the protection of personal information. A criminal can not only use the data they find to impersonate a banking customer but can also sell that information to others on the dark web for a handful of dollars per record. Anti-fraud has therefore become a battle between FSI entities and their criminal adversaries over who has the best data for training their AI models. Remember that a real-time payment transaction is not subject to rollback, so there is now more pressure to build models capable of making the right decision in real time.

Front of house vs. backstage

The data requirements mean that global payments companies that have been operating for decades will have a significant advantage. The more established a payments company is, the more data it can leverage through its payments orchestration platform in the fight against fraud. When sitting on billions of historic transactions worth trillions of dollars, an institution is in a better position to combine data depth with breadth and apply sophisticated AI and machine learning (ML) to spot patterns and adapt/evolve. Such an organisation is better able to spot anomalous behaviour using methods such as behavioural biometrics, where AI examines the mouse movements and typing and touch patterns of a user.

Behavioural biometrics allows continuous authentication of users without any extra security demands being placed upon them, thus providing a seamless frictionless payment experience. In a region where slick digital experiences are increasingly regarded as the minimum requirement of excellent customer service, banks and other FSIs gain a lot through AI. Front of house, they are delighting customers; backstage, they are beating the fraudsters at their own game. With enough high-quality data and the right AI tools, financial institutions can go beyond real-time fraud prevention and anticipate future criminal campaigns.

One hardly needs to draw a graph to visualise the drop off in effectiveness of AI-based anti-fraud systems as the scale and available budget of the bank decreases. Small and mid-sized banks are going to be vulnerable unless they partner with global payments processors that have the scale of data and the sophistication in AI to be effective against criminal elements.

Central payment infrastructures have a vested interest in protecting the payment rails. As the central point of transaction processing, they must ensure security to maintain trust, prevent fraud, and support the overall stability of the financial system. Payment companies have a stake in preventing fraud, too, and to that end will be only too happy to partner with any scale of bank to see the job done.

As time goes on, and the right partnerships form, all institutions will gain much more than the means to protect their customers from fraud. They will also gain monetizable customer insights. Greater volumes and higher accuracy of data will help banks identify macroeconomic and market trends, paving the way to new business opportunities, new products, and new services.

Safe at last

The intersection of AI and real-time payments in the Middle East represents a dual-edged sword — ushering in unprecedented convenience and economic growth while simultaneously amplifying the risk of fraud. As the region continues to innovate and expand its digital horizons, a collaborative and proactive approach to fraud prevention will be essential. Together, with the right strategies and partnerships, the Middle East can build a secure and prosperous future, leaving fraudsters with no place to hide.