UAB’s Braintrust Adds Artificial Intelligence to Fraud Fight

Bank Technology News | May, 2010

At United American Bank, the fight against anti-money laundering and fraud was becoming a tough job to handle through non-technological risk assessments—mostly handled by a single executive—so it’s turned to an automated behavior-based approach to locate suspicious transactions.

“I had some help from the branch managers who were looking at daily reports for suspicious activity, but it’s become a very labor intensive job,” says Joe Simoni, a vp at the UAB who handles the bank’s AML and anti fraud operations.

The $430 million asset bank has licensed anti-money laundering and fraud detection software form Verafin that uses a behavior-based approach that relies on artificial intelligence to seek out unusual behaviors. The software analyzes transactions and looks for patterns of behavior that are unusual for the customer or are indicative of money laundering. Simoni says the bank expects to deploy the technology by July.

The analysis creates a risk score for each customer and when the risk exceeds a certain level, the bank is alerted—along with all relevant transactions for the customers and the specific pieces of evidence that led to the alert.

“It allows [the bank] to grow from a manual environment with some rudimentary reports, but not a lot of technology to speak of,” says Andrew King, vp of client services for Verafin, which has 500 clients.