LONDON- Exchange group Nasdaq said on Wednesday that it is launching AI technology to help retail and commercial banks automate anti-money laundering (AML)investigations, as it expands into the financial crime software market.

The company hopes the system can make it quicker and cheaper for banks and other financial institutions to sift through the deluge of alerts flagging possible cases of money laundering generated by bank transaction monitoring systems.

The process of investigating alerts, which can potentially be as many as 300,000 a month, is currently manual, making it costly and labour intensive, Nasdaq said. Banks usually cast a wide net to catch illicit activity so the vast majority of these alerts ends up being false.

"Banks are worried about that wide net because of the cost," said Darren Innes, head of AML technology, sell-side and buy-side solutions at Nasdaq. "We are giving them the opportunity to reduce that cost." Banks have seen the number of alerts surge during COVID-19 as illicit activity attempts increased, Innes said.

While Nasdaq has long been a provider of market technology, including trade surveillance systems, the launch marks the exchange group's foray into the AML sector.

It comes as banks and other financial firms look to automate many of their more expensive and complex back office processes to reduce costs and increase efficiency.

"We have been thinking long and hard on how we want to go beyond trade surveillance," said Valerie Bannert-Thurner, senior vice president and head of sell-side and buy-side solutions, market technology at Nasdaq. "It's a product launch but strategically it's a launch beyond trade surveillance. We have great ambitions in the space."

Nasdaq's new system, which was built with UK-based startup Caspian, collates the data needed to conduct an AML investigation and analyses the information using software that replicates human decision making, the company said.

(Reporting by Anna Irrera; editing by David Evans) ((Anna.Irrera@thomsonreuters.com; +1 646 223 4005 ;))