Mastercard has recently introduced a groundbreaking approach to detecting compromised payment cards, utilizing the power of generative artificial intelligence (AI) combined with graph technology. This new method has led to a significant improvement, doubling Mastercard’s ability to detect compromised cards before they are used fraudulently.

This innovation is expected to have a wide-reaching impact on the global commerce ecosystem. Both online and physical retailers are likely to see a reduction in fraud-related losses, along with enhanced security for transactions. The new system could also result in fewer declined transactions due to suspected fraud, which in turn could boost sales and improve the overall customer experience. For eCommerce platforms, which constantly face the challenge of card-not-present fraud, this advanced technology could provide a much-needed defense layer.

“The new technology allows us to scan transactional data from billions of cards and millions of merchants faster than ever before,” explained Rohit Chauhan, Executive Vice President of AI-Fraud Solutions at Mastercard, in an interview with PYMNTS. He elaborated that this system alerts Mastercard to emerging, complex fraud patterns. “Using the generative AI techniques developed by Mastercard, we can extrapolate full card credentials from partially visible ones being sold on illegal online platforms. This means we are now able to double the rate at which we detect compromised cards, alert banks, and protect consumers before any fraudulent activity occurs,” Chauhan added.

Previously, Visa’s James Mirfin and Jotform’s Johannes Wiklund highlighted that small businesses are particularly vulnerable to fraud attacks. Additionally, PYMNTS reported that Mastercard’s First Party Trust Program, which employs AI and enhanced transaction insights to fight friendly fraud, is set to launch later this year.

According to a report by Juniper Research, online fraud continues to be a major issue in the financial sector, with projected global losses from payment fraud expected to reach $40.62 billion by 2027. Fraudsters employ numerous techniques to steal card data, including spyware, malware, and card skimming. These stolen credentials are often sold on illegal websites, where only partial card numbers are revealed to entice buyers while avoiding full identification.

The Role of AI in Fraud Prevention

Mastercard’s new fraud detection system integrates generative AI, which learns to create new data based on large datasets, with graph technology, which identifies patterns and connections among data points. This powerful combination enables the company to spot compromised cards before they can be used in fraudulent transactions, potentially saving banks and consumers millions of dollars.

The algorithm processes several factors, including recently reported fraudulent transactions, merchants known or suspected to be compromised, and pre-authorized transaction patterns. Rather than directly scanning illegal websites, Mastercard works with partners and third parties to collect the necessary data, ensuring that its fraud prevention efforts remain ethical and compliant with industry standards.

By leveraging AI, Mastercard claims it can now predict the complete 16-digit card numbers of compromised cards and assess the likelihood of those cards being used for fraud. This allows banks to block suspicious cards more quickly than ever, potentially preventing millions of dollars in fraud losses.

The system also creates a network of cards and merchants, generating links that reflect the associated risk. These connections are continuously updated with each new set of data, producing a dynamic list of at-risk cards that are likely being sold on illegal platforms. This adaptive system allows Mastercard to stay ahead of constantly evolving fraud techniques.

Mastercard’s graph technology, integrated with its AI-driven fraud detection system, provides more effective tracking across its network, which handles billions of transactions every year. For instance, the system might trace a single card to 200 others, identifying risky links to a merchant where compromised cards have been used. This detailed mapping allows Mastercard to issue faster and more accurate alerts to banks, enabling them to block and re-issue compromised cards swiftly, thus reducing the timeframe in which fraudsters can act.

The company has also embedded this advanced technology into its Cyber Secure product, which helps issuers and merchants assess their cyber risks and prevent potential breaches. Transactions on compromised cards are continuously monitored, ensuring that fraud attempts are swiftly detected, and cybersecurity measures are strengthened.

By implementing generative AI and graph technology, Mastercard is paving the way for a safer and more secure future for both consumers and businesses, setting a new standard in the fight against payment fraud.