How Banks are Winning the Battle Against Credit Card Fraud

In recent years, the landscape of credit card fraud prevention has undergone significant changes, driven by advancements in data analysis and machine learning.

by Faruk Imamovic
How Banks are Winning the Battle Against Credit Card Fraud
© Getty Images/Joe Raedle

In recent years, the landscape of credit card fraud prevention has undergone significant changes, driven by advancements in data analysis and machine learning. The Federal Trade Commission (FTC) receives thousands of card-fraud complaints annually. According to the Nilson Report, which monitors the card industry, payment-card fraud led to $33 billion in losses worldwide in 2022, with $13.6 billion of those losses occurring in the United States. Consequently, credit-card issuers and banks are increasingly focused on identifying and preventing fraudulent activities to safeguard their customers and minimize financial losses.

Historically, the decision to approve a transaction was based on factors such as the physical presence of the card, sufficient funds, and, at times, manual verification by cashiers. Today, the process has evolved significantly, leveraging data and advanced computational methods to detect fraud more accurately.

Modern fraud detection heavily relies on machine learning, a subset of artificial intelligence (AI). By analyzing extensive datasets, algorithms can identify patterns and relationships, creating decision trees to predict the likelihood of various outcomes and differentiate between normal and suspicious activities. Tina Eide, Executive Vice President of Global Fraud Risk at American Express, explains, "It's looking at what's happening that is very much out of the ordinary for your general behaviors. And when I talk about general behaviors, it is generalized, right? It is not down to the specific purchase or to the specific merchant." She adds, "The models are evaluating a trillion dollars' worth of transactions a year."

The Role of Data and Machine Learning

Mike Lemberger, Visa's Regional Risk Officer for North America, highlights the growth in data points generated by credit card transactions. "Visa, we don't have consumer information — that's your financial institution that has that — but what we have is this triangulation of all these data points. We can create more and better scores, layer on top of that machine learning and AI abilities, and it becomes a much, much more powerful predictor, which we then feed into all of our partners to say, 'Hey guys, if you want to make the best decisions, here's a whole bunch of really good information.'"

Visa, for example, does not directly block transactions but alerts banks to suspicious activity. Lemberger notes, "Visa can triangulate data points to create more accurate fraud scores. These scores are then shared with financial institutions to help them make informed decisions."

Yann-Aël Le Borgne and Gianluca Bontempi, researchers at the Université Libre de Bruxelles, emphasize the vast scale of fraud-detection technology. Companies process millions of transactions, generating numerous decision trees that can sometimes defy human logic. According to Bontempi, "Machines can consider many more features, and at the end of the day it's not clear if all those features have meaning for humans. Humans are used to working with two, three features, at most five features, while machines can work with hundreds of features. So there are really different levels between what a machine can do."

There are human-written rules, which are generally interpretable, and machine-written rules, which can be opaque. While machine-written rules are more accurate, they may be harder, if not impossible, to reverse engineer. Banks often use multiple algorithms, further complicating the process. Data scientists make the final decisions based on highly complex technology.

How Banks are Winning the Battle Against Credit Card Fraud
How Banks are Winning the Battle Against Credit Card Fraud© Getty Images/Justin Sullivan

Impact on Consumers and Privacy Concerns

Credit-card companies and banks continue to improve their fraud detection capabilities, despite the challenges posed by increasingly sophisticated fraudsters. The major credit-card issuers and banks in the U.S. generally have a zero-liability policy, meaning the organization, not the customer, absorbs the cost of fraud. This policy underscores the importance of effective fraud detection for financial institutions.

While some consumers may feel uneasy about the extensive data banks have on their spending habits, the benefits of enhanced fraud protection are clear. The advancements in machine learning and data analysis are making it more difficult for fraudsters to succeed, ultimately providing greater security for consumers.

However, the intricate nature of machine learning models and the sheer volume of data they analyze raise valid concerns about privacy. These systems can create detailed profiles of users' spending habits, sometimes knowing more about consumers' financial behaviors than the consumers themselves. Despite these concerns, the trade-off often leans towards enhanced security and convenience.

Tina Eide of American Express points out that the technology's effectiveness relies on understanding generalized behaviors rather than specifics. "It's looking at what's happening that is very much out of the ordinary for your general behaviors," she says. "It is generalized, not down to the specific purchase or merchant."

Mike Lemberger adds that the multitude of data points helps build comprehensive personas for fraud detection. "All these data points build personas. Just like in marketing, where personas are used to target ads, we're using that same technology to protect you," he explains. "These data points not only secure individual users but also enhance the security of the entire ecosystem."

Future Directions in Fraud Detection

As credit-card companies and banks refine their technologies, the ongoing battle against fraud is seeing significant progress. These efforts are crucial in a world where fraudsters continually evolve their tactics. Despite potential concerns about data privacy, the improved accuracy and efficiency in fraud detection represent a positive development in the financial sector.

Looking ahead, the integration of more sophisticated AI tools and real-time data analysis is likely to further enhance fraud detection capabilities. The aim is to stay a step ahead of fraudsters, continually adapting to new schemes and tactics. Financial institutions are investing heavily in technology to ensure they can protect their customers' assets and personal information effectively.

Additionally, the collaboration between financial institutions and technology companies is expected to grow. Sharing insights and advancements can lead to more robust systems that benefit all parties involved. This collective effort will be crucial in maintaining trust and security in the increasingly digital world of finance.