SINGAPORE - Citi has launched its payment outlier detection solution in 90 countries, including Singapore, which allows clients to review and approve payments which deviate from their past patterns of payment activity.

The solution, created by Citi's treasury and trade solutions business, identifies these outlier payments through the use of advanced analytics, artificial intelligence and machine learning. It was developed in the bank's Dublin innovation lab with direct client collaboration.

The solution uses advanced statistical machine learning algorithms instead of legacy rules-based logic to analyse payment patterns, Citi said, adding that clients do not have to opt in to the service.

It added that the launch follows an "extensive global pilot" with 20 clients such as Xerox Corporation and Tetra Lavel Corporation, with the service requiring minimum integration effort for clients.

The service will allow clients to have enhanced control and monitoring over their payments, the potential to reduce risk associated with outlier payments and subsequent losses, tailored customer profiles identifying payment patterns and real time alerts, to name a few.

Citi says the solution comes at a time where cyberattacks are becoming increasingly frequent and sophisticated. Therefore, banks need to be more vigilant than ever, given increasing transaction volumes.

The banking group is also leveraging its partnership with data science company Feedzai, which is backed by Citi's venture arm, to integrate Feedzai's transaction monitoring platform to provide enhanced service capabilities for clients.