Machine Learning is a prevalent method within the banking sector to address fraud. However, the results of the Machine Learning models, within certain cases are so cryptic they can be regarded as a “black box” – hardly anyone can explain precisely why specific outcomes are achieved.

Transparency is a crucial topic with data policy steadily growing in importance in Europe. To safeguard privacy laws and non-discriminatory modelling, regulations by the EU could consequently include the conditional use of explainable or interpretable AI. Therefore, we believe that a transparent approach to Machine Learning will become essential in how banks and payment service providers address AML operations.

We have discussed this with Roy Prayikulam, Senior Vice President of Risk & Fraud at INFORM – a global provider of solutions for AML compliance and fraud prevention, offering a flexible Financial Crime Prevention platform that enables customers to implement customer and transaction behavior monitoring in compliance with national and international regulations.