How Mastercard is Using AI to Enhance its Payment Systems
Mastercard, a global financial corporation, joins the AI race with its own generative artificial intelligence model designed to fend off scammers and enhance online shopping efficiency.
Mastercard connects over 22,000 financial institutions across more than two hundred countries. In 2022, the company's turnover surpassed $22 billion, with about 125 billion payments processed annually through the Mastercard network. This vast amount of data was used to train an artificial intelligence model aimed at safeguarding against fraudulent transactions, named Decision Intelligence Pro.
At the core of this innovation is a patented recurrent neural network architecture (such networks enable neurons to exchange data, processing both new information and insights from the network's previous state). The cybersecurity and fraud prevention division developed it from the ground up, incorporating open-source code where necessary.
Decision Intelligence Pro compiles a comprehensive history of clients' interactions with various goods and service providers. Leveraging this data, the AI predicts cardholder behavior patterns. When a new transaction is initiated, Decision Intelligence Pro evaluates its alignment with the client's usual spending habits. Alerts are raised for purchases that deviate from established patterns, with the analysis process taking just 50 milliseconds.
The company asserts that this cutting-edge technology for transaction decision-making can enhance fraud detection rates by an average of 20% across financial institutions, with potential increases of up to 300% in some cases. Furthermore, Mastercard anticipates that Decision Intelligence Pro could save at least one-fifth of the expenses banks typically incur for fraud detection and assessment.
An additional key feature of the model is its ability to preemptively identify fraudulent schemes. Through analyzing Mastercard client transactions, the AI uncovers emerging fraud techniques that could potentially spread.
This system is based on Dynamic Yield, a digital content personalization platform acquired from McDonald’s in 2022. Mastercard believes Shopping Muse will not only enhance user shopping experiences but also increase average transaction values while reducing return rates.
At the core of this innovation is a patented recurrent neural network architecture (such networks enable neurons to exchange data, processing both new information and insights from the network's previous state). The cybersecurity and fraud prevention division developed it from the ground up, incorporating open-source code where necessary.
Decision Intelligence Pro compiles a comprehensive history of clients' interactions with various goods and service providers. Leveraging this data, the AI predicts cardholder behavior patterns. When a new transaction is initiated, Decision Intelligence Pro evaluates its alignment with the client's usual spending habits. Alerts are raised for purchases that deviate from established patterns, with the analysis process taking just 50 milliseconds.
The company asserts that this cutting-edge technology for transaction decision-making can enhance fraud detection rates by an average of 20% across financial institutions, with potential increases of up to 300% in some cases. Furthermore, Mastercard anticipates that Decision Intelligence Pro could save at least one-fifth of the expenses banks typically incur for fraud detection and assessment.
An additional key feature of the model is its ability to preemptively identify fraudulent schemes. Through analyzing Mastercard client transactions, the AI uncovers emerging fraud techniques that could potentially spread.
The beauty of Mastercard’s ecosystem is we see data from all our customers globally from these transactions. What that does its it helps us actually see fraud and patterns across the ecosystem globally,explains Ajay Bhalla, Mastercard’s president of cyber and intelligence business unit.
Mastercard also introduced Shopping Muse, an AI-driven chatbot for online retailers, designed to tailor product recommendations for Mastercard cardholders. The bot makes recommendations by analyzing purchase history, online queries, regional specifics, and other factors. If it concerns buying clothes, the AI analyzes consumer tastes, and interest in various fashion trends, and can suggest items for creating a stylish look.
This system is based on Dynamic Yield, a digital content personalization platform acquired from McDonald’s in 2022. Mastercard believes Shopping Muse will not only enhance user shopping experiences but also increase average transaction values while reducing return rates.
Mastercard’s Chatbot Shopping Muse Source: neurohive.io
To wrap up, it's noteworthy that Mastercard has invested approximately $7 billion in artificial intelligence technology and cybersecurity developments over the past five years.
