Most Know Your Customer (KYC) processes utilized by financial institutions are inefficient, prone to errors, make for a bad customer experience, and cost institutions millions of dollars every year. Some of the solutions to help institutions become more effective and cost efficient include; robotic process automation, machine learning, and artificial intelligence. The focus of this article will be on Robotic Process Automation (RPA) in KYC, which really in its simplest terms is taking manual routine processes and automating them saving institutions money while being more efficient and effective.
KYC is a critical function to assess and monitor customer risk by establishing the identity of the customer, understanding the nature of the customer’s activities, and assessing the money laundering or terrorist financing risks associated with the customer. The manner in which KYC activities are completed and managed at institutions have an impact on client retention, the cost of compliance, and therefore the bottom line for the institution. The industry standard, up to this point, has been for every regulatory requirement change is to add additional staff to complete the additional workload to meet the new regulatory requirements and continuing to complete the legacy requirements in the same manner they have been completed for many years. Adding more manual processes on top of other manual processes increases the likelihood of errors, which leads to non-compliance of regulations and delays in onboarding potential customers.
Are these manual tasks really the expectation of the regulators? At the end of the day regulators are going to come in and assess the effectiveness of the program to be compliant with the regulatory requirements. If banks are able to develop processes that are more effective with minimal errors, standardized process, and free up those with subject matter knowledge to truly focus on identifying money laundering and terrorist financing risk and activities, I believe regulators would not only be ok with it, they would encourage it. Robotics isn’t going to take away the job of KYC professionals, since an effective robot will just complete the routine tasks and help streamline workflows, so the focus of skilled KYC workers can be on tasks requiring more specialized skills, critical thinking, and oversight. Everyone wants to catch the bad actors and there is a skill set required to do this that is becoming harder and harder to find and hire. So, if you can complete the repetitive tasks with a high level of confidence using RPA, freeing up time for the skilled workers to focus on identifying the bad apples, it seems like an option worth at least considering.
So, what are the main processes that RPA can really add value to the KYC programs at institutions? Less complex, repeatable, and stable processes that have high volumes tend to be the best candidates for RPA. Therefore, the main focus of RPA in KYC has been around entering and verifying Customer Identification Program (Name, Date of Birth for individuals, Address, and Identification number) data into the banks systems, compiling customer information from multiple sources and organizing it, plus customer screening and monitoring performing a first level review helping to eliminate “false positives” based on predetermined rules. This is completed when the robots interact with the institutions internal systems and external databases or websites/social media. Some of these databases have been used for years, such as sanctions lists, credit bureaus, the post office, and many of the KYC databases offered by third-parties. Most recently some of the largest banks in the United States have begun working with the DMVs and other governmental offices trying to explore new ways to digitally verify identity, since recent regulatory changes allow banks to use digital images of driver’s licenses in verify identity.
RPA is really about, at a minimum, mimicking what an analyst would gather, but with the ability to go to as many databases as the bank provides to find and organize the data on a customer. Than continue utilizing these databases on an ongoing basis to monitoring the institutions customers based on predetermined rules the business sets. This frees up the skilled workers to spend their time analyzing the data, instead of gathering it.
Another thing to consider is, if you have a large group of KYC analysts completing KYC tasks you probably have a large amount of inconsistencies across teams and lines of business. Even when they are following the same work instructions, individuals still tend to complete tasks differently, have variations in workflows, and even the most experienced workers may unintentionally have errors in these manual processes. RPA will take the process you define and will complete it the same standardized way every time. Plus, RPA usually has a full audit trail, when current manual processes may have a limited or incomplete audit trail.
There are lots of reasons to consider RPA from a better client experience, enhanced accuracy of information, increased productivity of staff, and large cost savings to the institution. Most institutions are using RPA at some level or utilizing third-parties that use RPA and don’t even realize it. Taking the time to consciously and strategically analyze the options and benefits for your institutions to utilize RPA may open the doors for some big wins not only from a cost perspective, but one the prevention and detection on money laundering by having a more effective KYC and AML program.