Edited by Sergio Silvestri

Google has also decided to address anti-money laundering activities with its own AI tools. Google Cloud, in fact, has launched Anti Money Laundering AI (AML AI).

This initiative has sparked a debate: to what extent can an AI-based system replace human intervention in detecting money laundering activities? I personally believe that human oversight is essential to prevent further fueling unfounded Suspicious Transaction Reports, which ultimately overwhelm the control system.

OGoogle’s stated goal for AML AI is to help its customers reduce operational costs while improving the effectiveness of their anti-money laundering system.

AML AI uses proprietary machine learning technology, as well as Google Cloud technologies, such as Vertex A.I e big query.

Among Google Cloud's future goals is to increase employee productivity by reducing the time it takes for an analyst to investigate potential suspicious activity.

AML AI, Google says, will enable greater risk detection, reduced operational costs, improved governance, defensibility, and customer experience. This will be achieved by providing a consolidated customer risk score, generated by machine learning, as an alternative to rules-based transaction alerts.

The risk score is based on bank information, transaction patterns, online behavior and data KYCThis score identifies high-risk cases and customer groups.

Google Cloud ensures that the proposed solution is able to adapt to variations in the reference data, providing results more updated and accurate that improve the overall effectiveness of the program and its operational efficiency.

"Google is an AI pioneer and is now leveraging its tools, technologies, and expertise to solve one of the financial services industry's biggest and most costly challenges," said Thomas Kurian, CEO of Google Cloud. "With AML AI, we will help financial institutions more accurately and efficiently detect money laundering risks, while improving business operations and governance."

Among those who have used AML AI so far are HSBC, The Hongkong and Shanghai Banking Corporation, Bradesco (one of Brazil’s leading banks), and Lunar (a digital bank).

Jennifer Calvery, Group Head of Financial Crime Risk and Compliance at HSBC, emphasizes that “AML AI has improved our ability to detect anti-money laundering activity. Google models are already demonstrating the enormous potential of machine learning to transform financial crime neutralization solutions across the industry. By enhancing our monitoring system with the sophisticated AI-based solution, we've been able to improve the accuracy of our financial crime detection and reduce the volume of alerts, resulting in less investigation time spent chasing red herrings. We've also reduced the processing time required to analyze billions of transactions across millions of accounts from several weeks to just a few days.”

We look at this experience with interest, but the five senses and intuition that humans have are still indispensable for this type of verification.