Financial institutions are looking to snuff out fraudulent banking transactions in real time as consumers increasingly fall prey to online scammers. To combat this, KeyBank’s robotic process automation (RPA) technology can automatically block bots using bots of its own that may attempt to carry out fake transactions.

Consumers reported losing more than $5.8 billion to financial fraud last year, a 70% increase from 2020, according to the Federal Trade Commission.

The bank has enhanced its fraud-detection capabilities through its use of robotic process automation (RPA) to detect account or transaction inconsistencies in real time. RPA and bots enable transactions to be flagged and reported immediately, Michael Reynolds, business technology senior manager of service digitization at $139 billion KeyBank, said this week during the panel discussion “Automation to Detect and Stop Fraudulent Transactions” at Bank Automation Summit Fall 2022 in Seattle.

“For example, when disputing credit card transactions, your Mastercard has a variety of different codes,” Reynolds said. “Each of them has different specifications of what you provide as information and the detail that you jump into.”

Using online purchases of virtual items for context, Cleveland-based KeyBank can identify when a smaller transactions are attempted in a short time frame from a questionable source and then begin retrieving disputed funds for its customers, he said.

As of today, 38 states permit Key’s APIs to connect to customer’s systems to block fraudulent transactions, according to Reynolds.

“Being able to process and get a jumpstart on [fradulent transactions] at least 38 [states] is helpful,” he said. “There’s a variety of mechanisms on the back end that you have to pay attention to. So, obviously after notifying the clients, people are told what’s happening to their account, and that’s important to us.”