Predictive Intelligence: “Game-Changer” for Preventing Payments Fraud Attacks?
- Predictive Intelligence has emerged as a valuable tool against fraud
- Even with many new payment methods available, checks remain popular fraud vehicles
- Predictive Intelligence anticipates the risk of fraudulent payments
Payments Journal reports that Predictive Intelligence has emerged as a valuable tool for financial institutions in their efforts to mitigate fraud attacks on payments.
Unfortunately, faster payment systems have increased fraud by giving criminals many more opportunities and a much wider pool of potential victims. The types of fraud typically seen are ACH, account takeover, fake merchant, and, of course, check fraud.
Oddly enough, with all the new innovations in payments, checks remain popular fraud vehicles. In 2022, check fraud increased by 96% from the previous year. What’s more, the average check value has risen from $673 in 1990 ($1,602 in today’s value) to $2,652 last year.
Consumer checks are mostly swiped from the U.S. Postal Service system, after which they are frequently altered to make counterfeits. Particularly troubling is a check fraud scheme whereby thieves use universal keys to access mailboxes, steal checks, and later change the payee information as well as the dollar amounts.
Business checks are not faring well, either, especially since these carry considerably higher dollar amounts and are highly lucrative targets for fraudsters. Early Warning’s report cited findings from the Association for Financial Professionals indicating that 63% of organizations fell victim to check fraud in 2022.
A New Approach?
Losses from fraud have increased significantly in recent years, and financial institutions often bear the costs -- hurting their reputation and customer trust in addition to impacting the financial bottom line.
Predictive Intelligence is seen by many as an effective defense. Using machine learning on data from many financial institutions, Predictive Intelligence anticipates the risk of fraudulent payments, allowing potential fraud to be stopped before any losses occur. For instance, by analyzing shared data from over 2,500 institutions, it generates risk scores to help evaluate payments and block fraudulent ones upfront.
However, Predictive Intelligence has been in use for check fraud detection for the last few years -- also known as behavioral analytics. This is the evolution of transactional analysis, moving from analyzing just transactions to entire account behavior. There are many use cases for how predictive intelligence/behavioral analysis provide key defense for check fraud, including:
- Drop Accounts: Fraudsters open these accounts to deposit stolen/fraudulent checks. There are two ways drop accounts are established: The fraudster will open an account online with little verification, or a fraudster will coerce a victim through social media to allow access to their account. Once the accounts are acquired, the fraudster will then deposit multiple checks into the account and extract the funds as soon as possible. Predictive intelligence/behavioral analytics will analyze these accounts for this behavior and flag these accounts to put holds on payments or close them.
- Anomalous behavior/transactions: Predictive intelligence/behavioral analytics monitors accounts and, if a check is written to a unknown payee or written for an irregular amount, the transaction will be flagged and put on hold until the bank can verify the validity.
- Geolocation: Predictive Intelligence/behavioral analytics can analyze the accounts and the locations where checks are deposited. If a transaction is performed in a location that is not common -- i.e. checks written for an account are deposited in the Northeastern US, but a large check is deposited in California -- this is a transaction that would be flagged by the system.
However, predictive intelligence/behavioral analytics can only do so much to prevent and detect fraudulent checks. Fraudsters are sophisticated and understand how these types of systems work. They will adjust by having checks written for amounts that are not outside the typical amount, or find mules to deposit the checks in areas that are typical of the account. That is why financial institutions need to deploy image forensic AI along with predictive intelligence/behavioral analytics. By analyzing both the behavior and the transaction itself, financial institutions build a stronger defense against check fraud.