"Check volumes are declining, but check fraud is going up". The first time I heard that statement was at a bank from their fraud specialist. It surprised me at first, but quickly made sense. Fraudsters are like water–they follow the path of least resistance.
Financial institutions are in a tough spot when it comes to identifying fraudulent checks. The float process, image-based check deposit channels, and rise of sophisticated image manipulation technologies enable fraudsters to attack in ways that are constantly evolving. Add to that mix the increasing prevalence and variety of scams and you have a recipe for rising fraud losses. There are two other sources of vulnerability within banks when it comes to fighting check fraud–aging technology and a lack of industry collaboration.
Outdated bank technology opens doors for fraudsters
While fraudster tools and attacks have become more sophisticated, the defense of banks, credit unions, and other financial institutions generally depends upon technology resources over a decade old (or worse). To borrow a phrase, banks are often“bringing a knife to a gunfight.” Consequently, fraudsters are winning an increasing amount of those fights. A 2020 report by the American Bankers Association reveals that attempted check fraud increased to $15.1 billion and accounted for 60% of attempted fraud against deposit accounts. The leading check fraud categories were counterfeit checks, forged signatures, and return deposited items.
Financial institutions have historically relied on data-centric fraud protection tools that often miss subtle (or obvious) visual clues that , if recognized, would identify suspicious checks or signal if they should be investigated.
Two major problems with old technology
Most payment fraud prevention tools used by banks today miss this. Another problem with old technology used by banks is that those check fraud solutions yield mountains of false positives that bank workers need to sift through. This is both inefficient and ineffective because bank reviewers fall victim to repetition bias, which means they can fall into a pattern of repeating recent choices. So, if the majority of checks are valid, a counterfeit check might be more difficult to spot and end up falling through the cracks.
For example, a group of reviewers within a bank that typically experiences a 1200:1 false positive rate – meaning they need to sift through 1200 items to find one bad item – are likely to miss signals that indicate possible fraud when a truly fraudulent check comes along. And with false positive rates like that, who can blame them? To find a similar environment with massive amounts of false positives, you need look no further than the nearest airport.
In our post 9/11 world, the Transportation Security Administration (TSA) is well aware of repetition bias. The organization employs a number of strategies to combat its effect. However, unlike the TSA - which must monitor every single item (bags, purses, etc) - a bank reviewer only needs to review selected items. Limiting that list to include only items that truly ought to be reviewed makes the entire operation more effective (and less costly).
The second area where banks are at a disadvantage is related to a lack of collaboration. Fraudsters collaborate and adapt much more quickly than banks. It is not unusual for us to see a check attack vector arise in the east coast in the morning and ripple across the US during the day as fraudsters share the latest vulnerability they’ve discovered. Banks, by contrast, might tap into an industry database (if they tap into anything outside of their walls at all). Most rely upon their own (old) tools to inform their actions, without accessing insights that would be otherwise available through collaboration. It’s not a fair fight.
Sophisticated fraud solutions for banks
In the modern era, artificial intelligence (AI), which leverages machine learning (ML) capabilities applied to large data sets, creates an environment where insights can be generated, tested, perfected, and applied much faster than yesterday’s software programs. AI/ML, when paired with the availability of high performance, secure cloud-computing services (which banks are increasingly relying upon today), open the doors to a new path that banks can take to significantly reduce check fraud.
These two technical insights, i.e. the ability to spot visual clues that indicate a bad check and other possible forms of fraud with technology, and the AI-enhanced collaboration advantages inherent in cloud computing, have enabled the innovative development of Mitek’s Check Fraud Defender, a network-based check fraud detection service that visually compares incoming checks against a profile of known “good” checks.
Mitek’s Check Fraud Defender detects forgeries otherwise missed by traditional fraud. The technology analyzes checks deposited from all channels – mobile deposit, in-branch and ATMs. With Check Fraud Defender’s patented computer vision technology, financial institutions have a higher degree of confidence in check fraud detection. Using a proprietary visual inspection model, Mitek leverages AI & machine learning to score 18 check attributes beyond account data. Overall, risk management is easier for banks because they are able to reduce fraud loss, centralize decision making, and minimize manual reviews.
Michael Diamond - Senior Vice-President and GM of Digital Banking at Mitek
Michael Diamond brought his talents to Mitek in 2012 and now, as Senior Vice-President and GM of Digital Banking, Mike defines and drives our Deposit Solutions unit, making it possible for more than 80 million consumers to deposit checks using their mobile phones.
Mike has held senior business development positions in global companies and curated strategic banking technology partnerships worldwide, as well as serving in Military Intelligence for the US Army Reserve, including deployment to Operation Desert Storm in 1991. Today, he focuses on expanding Mitek's reach into new areas of the financial services market, including mobile payments, digital deposits, and preventing check fraud. His zest for leadership, innovation, and learning has spanned 20 years of shaping winning cultures. he earned a Bachelor of Business Administration (BBA) in International Business from St. Norbert College. He's a well-known keynote speaker and regularly writes for his blog on leadership, innovation and personal effectiveness.