The development of new payment systems for consumers has inspired merchants, software vendors, and financial institutions to become more creative in combating fraud. Artificial Intelligence has emerged as the go-to solution for reducing risk. Next generation AI promises to be even more of a game-changer in the world of fraud detection, not just uncovering but also anticipating fraudulent transactions.
With the increasing growth of payments data, acquirers and merchants are finding it harder to get a comprehensive view of consumers’ behavioral patterns. This leads to a fragmented approach to fraud prevention, making it difficult to determine what is a legitimate transaction and what is fraud. Models trained on global data allow for a comprehensive view of consumer transactional patterns, resulting in increased fraud detection and approval rates with fewer false positives.
“Artificial intelligence allows us to protect the 125 billion transactions we switch on our network every year at speed and scale,” said Rohit Chauhan, Mastercard’s Executive Vice President of Artificial Intelligence. “By applying thousands of data points, our sophisticated AI engine helps banks approve more genuine transactions and prevent fraud. In fact, our AI-powered solutions have saved $35 billion in fraud in the past three years alone.”
Mastercard has been using AI for more than a decade, most importantly in its cybersecurity work. As part of Mastercard, Brighterion has developed AI fraud models that monitor transactions from all sides to ensure accuracy in predicting fraud. Its AI technology checks against multiple transaction indicators and compares them with patterns identified in historical fraud.
Introducing the Next Phase of AI
Mastercard has combined its AI and payment gateway capabilities to deliver a unified solution, Transaction Risk Management powered by Brighterion AI, that enables acquirers to proactively detect, prevent, and mitigate fraudulent activities. Transaction Risk Management leverages AI and machine learning technology to provide real-time analysis, enabling acquirers to use advanced technology to better protect their merchants. The result is an easy-to-use solution that can reduce fraud and approve legitimate transactions more effectively.
Through Transaction Risk Management, each transaction is evaluated in two paths—there’s an AI model and there are also the rules set by the customer. Firstly, The AI model checks against multiple transaction indicators and compares them with historical patterns as signals that are correlated with fraudulent use. AI keeps a continuous eye on the model to evaluate when adjustments might be necessary.
The solutions second path assesses the transaction with a rules management tool. Customers can use a variety of rules within the supported templates, as well as establish their own based on business specifics. After the assessment, each transaction is assigned a numerical score that indicates the level of risk associated with it. When the two models are integrated, they give a clear assessment of when a transaction might be fraudulent.
The Value of Experience
Mastercard has a long history of embracing AI to secure the digital ecosystem. A primary focus is providing fraud detection and enterprise Al applications for payment service providers, financial institutions, healthcare payers, and merchants.
“Mastercard and Brighterion have substantial experience applying AI technology to fraud detection,” said Christopher Miller, Lead Analyst of Emerging Payments at Javelin Strategy & Research. “They have been using AI in fraud detection before many of the more recent AI entrants were even around. As part of Mastercard, Brighterion can distribute this technology to a much wider audience than then they could ever have achieved alone.”
Customers can leverage the expertise of Mastercard across a diverse skill set, and the payment strategy works alongside an end-to-end service that focuses not just on the technology but also on customer service and experience. Brighterion AI’s full-stack machine learning toolkit creates off-the-shelf market models that are production-ready, and custom models are available within six to eight weeks.
Existing Applications
The processes have already been put to use around the globe. Earlier this year, Mastercard announced a partnership with Network International, the leading enabler of digital commerce in the Middle East and Africa, to address fraud, declines, and chargebacks while reducing costs and risks for acquirers. Leveraging Mastercard’s Brighterion AI technology, Network International expects to provide transaction fraud screening and merchant monitoring to its customers across the region.
“At Mastercard, we think of AI like electricity: powering our society, enlightening our communities, and driving progress,” Chauhan said. “That’s why we use it everywhere we can.”