Legacy check fraud detection tools have a very high-false positive rate — shockingly, there’s usually only one fraudulent check for every 800 that get flagged for manual review. This puts a lot of strain on your employees and by extension on your bottom line, but it doesn’t have to be that way.
By combining legacy check fraud detection tools and check image analysis systems, banks can reduce their false-positive rates. Some analysts say that it can get as low as one in 30. Even more importantly, this approach can detect fraud more efficiently while reducing the need for expensive human input. To learn more, contact us today, or keep reading for an overview.
Rule-Based Legacy Check Fraud Detection
A lot of legacy check fraud detection relies on rules. If checks are outside of the set parameters, they are flagged for manual review. For instance, rules often address checks over a certain amount or checks written on new accounts. This creates a lot of false positives, but it also opens the door for fraud. As soon as the bad actors realize what amount the bank flags as suspicious, they can just start targeting checks under that value.
Forgery Analysis
To spot forgeries, banks historically relied on manual analysis, but of course, they didn’t look at every check. They usually only looked at checks that had been flagged by the rule-based system, or they relied on tellers to spot aberrations. Unfortunately, even a well-trained and very experienced signature analyst often can’t tell the difference between a skilled forgery and a real signature.
However, check image analysis tools can spot forgeries very effectively. They look at both static and dynamic elements of the signature. For instance, they look at the shape of the signature, but they also take into account the fluidity of the signature.
Signature verification tools can also assess multiple reference signatures, and this improves their accuracy. In contrast, human analysts tend to have reduced accuracy when trying to detect a forgery using multiple reference signatures.
Customer Verification
In the past, bankers often looked into suspected fraud by contacting the customer. They would call the customer and ask if they deposited a check in a certain amount, and then, if the customer verified the check, they would let it clear. However, this doesn’t work in the face of new account fraud.
With new account fraud, the check may be legitimate but the account holder is not. They’ve either opened the account with a fake identity, a stolen identity, or a combination of made-up and stolen details. If the bank calls to verify, the account holder will say the check is fine. But then, they may overdraw their account and run or finish whatever other type of fraud they were planning to do with the account.
MICR and Challenges With Mobile Deposit
Magnetic Ink Character Recognition (MICR) has long been a staple in check fraud detection. Checks must be printed with their routing and transit numbers in magnetic ink. This allows the banks to scan the checks automatically. Historically, it also helped to prevent fraud because criminals couldn’t easily print fake checks if they didn’t have a MICR printer.
The Federal Reserve still requires checks to have MICR data, but now, thieves can get around this requirement by using mobile deposits. When you take a photo of your check for mobile deposit, the app can’t tell if there’s magnetic ink on the check or not. As a result, you must look at different signals when trying to assess if a check is legitimate or not.
In particular, you should have tools that help you verify the identity of the customer making the deposit. This should include their sign-in and password, but your system should also look at their IP address and any recent actions that could indicate account takeover.
Fraud Detection Software and Hardware
Legacy check fraud detection relied on memory-intensive software with expensive hardware needs. This was very costly for banks to implement, especially if they were small community banks or credit unions with modest budgets.
However, modern systems also offer a way to mitigate this issue. When you opt for cloud-based fraud detection along with hosting, you don’t have to worry about making a big upfront investment into software and hardware. Instead, you can subscribe to the services you need and also get support from your managed fraud detection partner.
Improving Check Fraud Detection
To improve check fraud detection, you need to use the legacy tools noted above, but you need to improve them by integrating check image analysis. The integration should involve the following:
- Risk and rules-based engines — You can’t only rely on rule-based check fraud detection. You need a system that can also assess transactions based on their risk. Generally, this is established by a system that can learn about fraud patterns and customer behavior.
- Artificial intelligence — AI plays an instrumental role in fraud detection by looking for fraudulent patterns that signal forged checks, stolen checks, account takeover, and new account fraud.
- Cloud-based tools — This helps to ensure that you don’t face any bottlenecks due to subpar equipment or excessive upfront costs for hardware or software. Additionally, the cloud allows you to get managed fraud detection services where third parties review suspicious items during the weekends or evenings when your fraud team isn’t working.
You also need a solution that will help your financial institution look for check fraud across multiple channels. This includes mobile deposits, ATM deposits, branch deposits, and check clearing houses.
At SQN Banking Systems, we’ve been in the industry for decades so we understand legacy fraud detection tools. We also understand the critical importance of evolving if you want to protect your bank. To learn more and to talk about how to protect your bank from check fraud, contact us today.