Which Company's Case Study?
This is a case study of Mastercard, a global payment technology company. To address the long-standing challenge of credit card fraud, Mastercard built 'Decision Intelligence,' an AI-powered system that assesses transaction risk in real time, and offers it to financial institutions that issue cards.
Challenges They Wanted to Solve
With the expansion of e-commerce, credit card fraud has become more sophisticated, and methods have diversified. Traditional rule-based fraud detection systems faced challenges such as slow response to new fraud techniques and 'false positives' (incorrectly blocking legitimate transactions) that damaged the customer experience.
How AI Was Used
Mastercard leverages an AI platform called 'Decision Intelligence.' This system analyzes 143 billion transactions annually using machine learning to score the likelihood of individual transactions being fraudulent in real time. Card issuers can use this score to instantly decide whether to approve, request additional authentication, or decline a transaction. Furthermore, 'Decision Intelligence Pro,' which incorporates new generative AI technology, scans a trillion data points and evaluates relationships between entities to further improve detection accuracy.
Introduction Effects and Key Takeaways
- ▸It has been reported that the introduction of generative AI technology has improved fraud detection rates by an average of 20%, and in some cases, up to 300%.
- ▸Cases of legitimate transactions being incorrectly rejected (false positives) have been reduced by over 85%, contributing to increased customer satisfaction and merchant sales.
- ▸The key takeaway is that it not only detects fraud but also significantly enhances the customer experience by 'reducing false positives.' This is an excellent example of using AI to balance often conflicting challenges: strengthening security and improving convenience.
What Japanese Companies Can Learn
In industries that handle vast amounts of transaction data, such as finance, retail, and telecommunications, similar approaches can be used for real-time risk assessment and fraud detection. Rather than developing everything in-house, utilizing AI services offered by companies like Mastercard via APIs can be an effective option for rapid implementation. The sense of balance in strengthening security without compromising the customer experience will provide important insights for many companies.