Banks are turning to robotic process automation (RPA) to automate collections, make core transitions and help with fraud and anti-money laundering compliance (AML) this year.

Banks are automating loan collections because they’re anticipating a wave of loan defaults as COVID-related restrictions on collections expire, said Amit Kumar, vice president of financial services at RPA provider UiPath, rated as one a leading RPA vendor by IT research firms Gartner and Forrester.

“Banks really need to be proactive now, because they are seeing this problem on the horizon,” Kumar told Bank Automation News.

Banks are using technology to automate loan collections in three ways, Kumar said. First, they are using artificial intelligence (AI) technology to predict the percentage of clients — and which specifically — are at risk of default. Second, banks use RPA to reach out to customers to manage repayment or restructuring options before a loan goes into forbearance. Finally, banks are leveraging automation technology to quickly provide call center agents with all relevant loan information when a customer calls.

Banks are also showing more interest in using RPA to address AML, Kumar said. RPA can be used to reduce the false positives, a major concern for compliance officers,  as well as to collect data on fraud alerts that need further investigation.

“I would say we’re also beginning to see how AI can jumpstart the investigation,” Kumar said, adding that this is especially true when looking at unstructured data like news articles and contracts.

One reason more banks may be deploying RPA in this way could be that vendors are starting to embed machine learning models for fraud detection and AML, according to IT research firm Gartner. In its November 2020 report on emerging trends in RPA software, Gartner noted that new use cases would evolve this year based on the following three enhancements to RPA platforms:

1. Embedded machine learning (ML) models;
2. Integration of process discovery and process mining; and
3. Cloud delivery.

“An emerging trend is for RPA software vendors to offer prebuilt ML models designed for specific use cases and workflows,” the Gartner report noted. Other examples include loan originations for financial institutions and claims processing in insurance.

Another use case on the rise is converting from a legacy core to a new core, said Karen Reichle, the vice president of customer success engagement at workflow management and RPA software company Nintex. Banks turn to RPA tools primarily to cut the time down for the core conversions. Core vendors often have a backlog of 10-12 months, Reichle said, and RPA can be used to move customer records and account numbers over to the new system in about 12 weeks.

Another use case Reichle has seen grow this year is the use of RPA to process information in payments systems. Bots can listen to new workflows that come in and automate any change. For instance, if a customer is moving $25 per month into savings and wants to change that to $100, the bot can fulfill the request.

“You gain a lot of efficiencies, but you also gain accuracy when using a bot” to automate such transactions, Reichle added.