Mastercard's generative AI, an asset in the fight against fraud

Mastercard reports on the evolution of its anti-fraud system, highlighting the results of using generative AI to detect cards whose data may have been compromised. An effective way of reducing risk, at a time when fraud is on the rise in sectors such as retail.

FACTS

• Mastercard recently announced the evolution of its generative AI-powered fraud detection systems.

• These developments are aimed in particular at card frauds such as “card skimming”, which are systems devised by fraudsters to recover cardholder data (16-digit codes, PIN codes, expiry dates, but also personal data).

• In concrete terms, these procedures enable fraudsters to extract this information in physical situations, such as from ATMs, or online, when making a payment on a website infected with malware.

• To counter these data hacks, Mastercard uses generative AI to predict which cards have been compromised.

• To achieve this, the technology analyzes large quantities of card and merchant transactions. Thanks to this, the system identifies fraud and can then anticipate and therefore protect threatened transactions. This includes :

• Twice as effective detection of compromised cards;

• Nearly 200% fewer false positives when detecting fraudulent transactions on potentially compromised cards;

• 300% faster identification of merchants at risk or compromised.

• Mastercard can therefore warn banks affected by card fraud more quickly. They can then alert card owners to block compromised cards.

CHALLENGES

• Strengthening the bond of trust with financial institutions: With its solution, Mastercard reinforces its position as one of the leaders in cybersecurity in the banking system. The significant rise in fraud detection scores and the company's technological advances make it a major partner for banks wishing to protect their customers. By communicating these results, Mastercard is strengthening its bond of trust with the various players in the banking world.

• Results that confirm the effectiveness of generative AI: last February, Mastercard integrated Decision Intelligence Pro into its system, a generative AI model designed to reduce and prevent fraudulent transactions. The results communicated by the company show a significant improvement in the system reinforced by generative AI, particularly in fraud prevention, a key point in the fight against fraud.

• The fight against fraud, a major focus of Mastercard's strategy: The scheme has made cybersecurity one of the pillars of its current offering. It relies in particular on RiskRecon, a company acquired in 2020, and Cyber Secure, the part dedicated to cyber security and anti-fraud activities. Aware of the opportunities offered by generative AI on these subjects, the company has rapidly integrated the technology into its solutions, enabling them to considerably increase their effectiveness.

MARKET PERSPECTIVE

• The fight against fraud is a major issue for all payment networks. In early May, Visa announced an update to its Visa Account Attack Intelligence (VAAI) tool, including the integration of a scoring system powered by generative AI, to help prevent fraud in cardless transactions.

• Adyen's annual retail survey shows an increase in fraudulent activity worldwide. For example, 35% of consumers surveyed were victims of payment fraud in 2023, compared with 23% in 2022. This sharply rising figure may explain schemes' massive investment in anti-fraud issues.