Leveraging AI and Machine Learning to Fight Financial Crime

Money laundering schemes were originally created by organized crime to “clean” money made during U.S. Prohibition in the 1930s. Law enforcement caught on and laws were passed to curb the illegal activity. But today technology has made the criminals more emboldened, more creative — and more successful.

The amount of money laundered annually is up to five percent of global GDP (up to US$2 trillion), according to the United Nations Office on Drugs and Crime — and it appears to be getting worse. The Basel Anti-Money Laundering Index, an independent ranking of 129 countries by the International Centre for Asset Recovery, found that there is little measurable progress in countering money laundering activities and that 42 percent of countries worsened their risk scores between 2017 and 2018.

It’s a trend that SAP partner Quantiply is acting to remedy. Today, the company is working closely with SAP on anti-money laundering (AML) and know your customer (KYC) solutions capable of identifying and analyzing suspicious financial transactions, said Surendra Reddy, founder and CEO of Quantiply.

“We believe that working together with FIs and Regulators we can stop financial crime. We’re well on our way to making this change happen. Quantiply is passionate about this endeavor and fortunately we have connected with like-minded people who want to achieve the same goal,” Reddy said.

Separating the Good Guys from the Bad Guys

Television shows and movies have helped put a spotlight on money laundering and other financial crimes. Governments have also stepped up efforts to thwart these schemes, in part because the funds are being used for even more nefarious crimes, according to John Tripier, vice president of Business Development and Strategy at Quantiply.

“There’s a shift happening to crack down on financial crime and there’s a lot of concern about the number of people doing bad things. We’re looking to empower banks and regulators to do a better job of finding and stopping money laundering that’s now being used not only by drug lords but also to fund terrorism, child pornography, human trafficking, and other horrible things,” Tripier said.

In the U.S., banks are required to report cash transactions of more than $10,000 to regulators agencies. The problem is that banks only know you made a large cash transaction, they don’t know if you’re a good guy or a bad guy, Tripier said.

Quantiply developed the Quantiply Sensemaker applications suite as an alternative to older manual or rules-based solutions. Sensemaker uses artificial intelligence (AI) and machine learning algorithms to identify potential bad actors and high-risk transactions, helping financial institutions and regulators flag suspicious activities with better accuracy. By reducing operational costs and risks, customers are more likely to monitor and seek penalties for financial crimes, Reddy said.

“Our job is to help regulators go after the bad guys. There needs to be a system built on continuous learning. AI can help develop trends based on customer activity timelines. With Sensemaker, a bank can monitor what the customer does in those interactions, who they have associations with, and whether they’ve had bad transactions in the past,” Reddy said.

Using AI to Protect Customers

Historically, up to 95 percent of flagged transactions generate false positives under rule-based approaches and it can take three to five hours to process each case. AI-based systems learn from customer activities and get stored on an SAP HANA platform that allow users to make real-time judgments about each case.

“Out of the box, we reduce false positives by up to 65 percent. But really, it’s not just about false positives, or negatives, but when you can prevent risk rather than cure the risk,” Reddy said. “Banks want customers to know that they will have a good financial experience. If you’re a bad guy, banks don’t want to do business with you. Our goal is to not only enable them to create better financial institutions, but reduce their risk with customers too. You don’t want your money in a bank that enables illegal activities.”

Quantiply Sensemaker users typically see value after 90 days, instead of two years under older solutions, Tripier said. The goal is to create a groundswell movement where financial institutions, regulators, and other potential customers recognize that financial crimes can be stopped in an affordable way.

“Our goal — our purpose — is to make change happen. We’re not using technology to solve world hunger, but to make changes to financial institutions and then society,” he said. “That’s our vision and our thinking. We’re getting traction. SAP and others are working very hard with us to solve these problems.”

SAP and our partners. Improving lives. That is our purpose.