Algorithmic Compliance under the Prevention of Money Laundering Act, 2002: Efficiency Gains, Liability Gaps, and the Need for Explainable AI

Abstract

The rapid digitalisation of financial systems, along with the multi-fold increase of the volumes of transactions, has forced the reporting entities to resort to the algorithmic means of the Anti-Money Laundering (AML) compliance even more. The Prevention of Money laundering act 2002 (PMLA) has committed the financial institutions to monitor suspicious transactions and maintain records as well as reporting any potential money laundering transactions. In this respect, compliance systems based on Artificial Intelligence will offer considerable efficiency benefits of scoring risks automatically, real-time tracking, and a more effective identification of patterns.