2  The Fight Against Fraud: How Companies are Using AI to Fight Back 

The Fight Against Fraud: How Companies are Using AI to Fight Back

With Technology evolving so immensely, fraud is also becoming a bigger risk to people through Technology. Fraud is way more common today due to the advancement of technology and the number of people who have dealt with fraud continues to rise. The Federal Trade Commission reported that “Consumers reported losing more than $3.3 billion to fraud in 2020, up from $1.8 billion in 2019. Nearly $1.2 billion of losses reported last year were due to imposter scams, while online shopping accounted for about $246 million in reported losses from consumers”. The amount of money people have lost to fraud has gone up almost every year, and how are we going to stop it from happening? With technology constantly advancing and fraud on the rise, many companies like Mastercard, are implementing new ways to use AI to prevent fraud with the use of Machine Learning and the three ways MasterCard is using AI. The ways Mastercard uses AI and Machine Learning can be very useful for customers and for businesses to gain their trust.

Many people within the past decade have been worried about their data getting leaked, hacked, or taken online. They never know how to prevent it, whether businesses are taking advantage of them, or they are getting tracked online. But with technology advancing, so is AI Fraud prevention, “Machine learning is a set of methods and techniques that let computers recognize patterns and trends and generate predictions based on those” (Ravelin.com). Machine Learning is used within our everyday lives, whether it is used for detecting spam emails, recommending certain products, and used in the banking industry to find unusual transactions. In the process of Machine Learning, it takes in large amounts of data/information and learns how to make certain decisions on the data. Machine Learning is perfect for fraud detection because it is more efficient, faster, and more accurate than humans who would be detecting patterns and transactions, the cost of doing this is considerably lower. This brings up the question of why many companies have not integrated Machine Learning AI. Many consumers still are unable to trust AI, as the “AI Regulation Is Coming” from the Harvard Business Review explains “AI systems that produce biased results have been making headlines. One well-known example is Apple’s credit card algorithm, which has been accused of discriminating against women, triggering an investigation by New York’s Department of Financial Services” (Candelon 1). Because there have been troubles in the past with certain companies when using AI, consumers have trouble trusting AIs. People need to understand that AI’s make mistakes too, just like humans. Even if Machine Learning fails and makes one mistake, people should not be worried. From an article in the Harvard Business Review,  “AI Regulation Is Coming”, the authors explain how AI and Machine Learning are moving into the direction of AI explaining their decisions. With Machine Learning and new algorithms, it allows it to learn from its mistakes and fix what went wrong. Machine Learning is more efficient, accurate, and scalable to help fight against fraud.

As one of the largest credit card companies in the world, Mastercard faces fraud daily. They have changed their way of fighting against fraud by using AI because payments are now being made digitally. So many people use their technology to pay for items instead of paying with cash or even a credit/debit card. Phones can now store one’s credit card or debit card information and make paying much easier. But, this also makes stealing information and fraud easier. Mastercard implemented three ways of using AI to fight against fraud. The first way that Mastercard uses AI to fight against fraud is their In-memory grid, which helps facilitate fraud analytics. This AI tracks and finds irregular patterns and payment behaviors, “refreshing constantly to analyze new transactions against a cardholder’s purchase history” (Boulton 1). It is very similar to Machine Learning, being able to follow patterns and find irregular patterns in one’s payments. President of operations and technology at Mastercard, Ed McLaughlin, said “that the grid stopped $1 billion in financial fraud since 2016” (Boulton 1). The second way that Mastercard prevents fraud is by using AI “smart agents” to guard the front gate. This AI supports other AI in the system like the in-memory grid and machine learning. It generates risk models and scores to approve purchases made by MasterCards. Having AI help other AI is great because if one makes a mistake, the other can fix it. The last way Mastercard uses AI to go against fraud is their NuData security. The NuData security looks at how someone uses their own personal technology and looks for patterns and how frequent they use it. “NuData software assesses, scores and learns from each online or mobile transaction to enable merchants and issuers to make near real-time authorization decisions. Anything outside normal usage suggests aberrant behavior that could raise red flags” (Boulton 1). So if someone else is trying to use your MasterCard or on your account, the NuData can tell that it is someone else by the behavioral actions and lock them out. Mastercard uses AI effectively to prevent fraud from happening, and will continue to make the AI more effective as time goes on.

Lastly, I want to talk about a fraud that occurred in Indiana and how it could have been prevented if AIs were used. In this fraud, church members defrauded the state of Indiana of $62,000 in food stamps and welfare assistance. Because the pastor and his wife had an income of over $100,000, they should not have been able to get welfare assistance and a member of the church created fraudulent accounts for eleven different people, thus allowing them to use the benefit cards to draw funds or food stamps at ATMs and grocery stores. This fraudulent act took $62,000 away from the taxpayers in the state of Indiana. How would AI prevent this fraudulent act? Because this incident happened well over 10 years ago, efficient and accurate AI might not have been in place. From either the AI MasterCard uses or from Machine Learning, this incident could have been prevented by AI. Machine Learning would have recognized an irregular pattern, and unusual transactions. AI would have been able to red flag the eleven people that defrauded the state and took $62,000 from the state and taxpayers. If a system like NuData, that MasterCard uses, was set in place for the state of Indiana could have also prevented this. NuData uses behavioral analytics, habits, and expectations. If neither of these were met, they would have been red flagged. Because many frauds use fake names or are anonymous , they have a feature called Device Intelligence. Device Intelligence “gathers hundreds of data points from the device, network, connection, and location. The real-time assessment helps recognize devices seen in the broad NuData network, which monitors billions of events every month” (NuData Security). This feature would have tracked down where the fraud occurred, and the usage of the eleven fake welfare cards. If AI systems like the Machine Learning or NuData were in place during this food stamp and welfare assistance fraud, it could have been prevented.

In conclusion, many people see AI as either evil or good, but depending on how they are used they can be helpful. Fraud is such a big issue in our country, but also the world. Even though AI is not perfect, like us humans, they are still learning and evolving to become even more helpful in preventing frauds but also in many other things. Many people are losing a lot of their money due to hackers and frauds, because technology has advanced so much within the past decade AI will be able to fight back for us. MasterCard has started a plan to help prevent fraud with their use of AI, and hopefully many other companies will plan to help their customers as well. AIs are the future of preventing fraudulent acts and saving people’s money. As more and more companies begin to implement AI for their security and the AI continues to learn and grow, the number of cases of fraud will go down. AI like Machine Learning will be able to predict future fraud attacks/risks, so for instance the food stamp fraud that occurred in Indiana will not happen again because AI will be able to predict future fraud risks. If we cannot trust or let AI learn and evolve, then we could be at a bigger risk of more fraud to happen. The future is set to have the world revolve around technology, so why not have AI help us be protected?

APA Works Cited

Candelon, F., Charme di Carlo, R., De Bondt, M., & Evgeniou, T. (2021, August 30). AI Regulation Is Coming. Harvard Business Review. Retrieved November 14, 2021, from https://hbr.org/2021/09/ai-regulation-is-coming.

Boulton, C. (2018, December 3). 3 Ways MasterCard Uses AI To Fight Fraud. CIO. Retrieved November 14, 2021, from https://www.cio.com/article/3322927/3-ways-mastercard-uses-ai-to-fight-fraud.html.

Wthr.com. (2006, May 9). Church Leaders Charged With Food Stamp Fraud. wthr.com. Retrieved November 14, 2021, from https://www.wthr.com/article/news/church-leaders-charged-with-food-stamp-fraud/531-4351d507-4cf0-42d0-91ca-32ea302abb17.

Machine Learning For Fraud Detection. Ravelin. Retrieved November 14, 2021, from https://www.ravelin.com/insights/machine-learning-for-fraud-detection.

How It Works – NuData Security – How The Technology Works. NuData Security. (2021, March 30). Retrieved November 14, 2021, from https://nudatasecurity.com/how-it-works/.