A Fraud Series - Part Three: Types of Fraud Identified by a Detection System

In the first post of this series, we quickly reviewed some of the many types of fraud that have occurred over the years. In this post, we will discuss these types of fraud in more depth and provide examples:

Chargebacks: Charges made to a financial account that were not authorized may result in a chargeback. For instance, a compromised credit card can be used to purchase goods without the card holder’s knowledge. If this transaction is successfully disputed, the money can be returned to the card holder’s account. The damage, however, has already been dealt. Oftentimes the fraudster has made off with the illegally purchased goods while the businesses are held responsible for returning the stolen money to the card holder. In some cases, chargeback fraud can be committed by the card holder themselves. In this case, the card holder purchases an item, receives the item, and then disputes the transaction with the bank to initiate a chargeback. This pattern is extremely common on the internet. 

Illegal Offshore Accounts: Using offshore accounts is not necessarily illegal. There are many legal reasons why someone would want to move money into an offshore account. Moving money to offshore accounts in order to evade taxes or hide assets, including assets that were acquired through criminal activity, however, is illegal. When analyzing account interactions, it can be difficult to see and follow the links between various entities and the accounts connected to them. Graph analytics help fraud analysts visualize connections between potentially fraudulent entities in an intuitive way that makes it much easier to follow fraud rings in a system. 

Money Laundering: Money laundering involves legitimizing illegally acquired funds. For example, fraudsters can set up business fronts to channel stolen money, making the money appear to have been legally sourced. One can also split large amounts of money into smaller transactions or even spread these transactions across multiple accounts to reduce suspicion. As you could imagine, with a little creativity, there are a ton of ways to clean dirty money. Fraudsters are using the latest technologies to hide their money under a layer of obscurity, performing transactions between many accounts, through legitimate sources such as restaurants or cryptocurrency, making it very difficult to trace the money back to its original source. This is partly why keeping up with fraud trends is so difficult for blacklist and rules based fraud detection systems. Systems built to adapt to these trends rather than react to them are much more valuable for capturing fraudulent or anomalous activity. Tune in next week to learn how we might surface money laundering in our Connected Toolkit.

In this post we’ve reviewed some common types of fraudulent behavior and accompanying examples. It is important to note that as time goes on, fraudsters will continue to adapt to rules and regulations that attempt to prevent them from committing fraud. While supervised detection systems can detect certain types of fraud well, oftentimes analysts don’t immediately know whether new patterns are indicative of fraud or not. This is why it is important to use an unsupervised system that can learn not only from existing patterns, but from new patterns as well.