Transaction monitoring in anti-money laundering: A qualitative analysis and points of view from industry
Berkan Oztas,Deniz Cetinkaya,Festus Adedoyin,Marcin Budka,Gokhan Aksu,Huseyin Dogan
DOI: https://doi.org/10.1016/j.future.2024.05.027
IF: 7.307
2024-05-20
Future Generation Computer Systems
Abstract:Financial institutions face significant challenges in their efforts to prevent money laundering and terrorist financing due to criminals' continuously evolving techniques and the vast volume of transactions that need to be processed. The traditional rules-based methods utilised in banks produce high false positive rates, which lead to increased costs and inefficiencies. This study identified the perspectives of 8 anti-money laundering (AML) specialists on the current state and potential improvements in transaction monitoring methods. The results provide in-depth knowledge of the problems and requirements for researchers and practitioners. Semi-structured interviews conducted with the AML experts (totalling 480 min) identified the challenges, requirements for successful implementation, and future trends in transaction monitoring. The findings reveal a growing interest in machine learning and artificial intelligence to enhance the efficiency and accuracy of current approaches. Furthermore, innovative methods such as graph analysis and anomaly detection were suggested to overcome the limitations of rule-based systems. Requirements such as explainability, flexibility, and identifying new risks were extracted and analysed. This research contributes to the existing literature by providing valuable insights from industry experts, guiding the development of advanced transaction monitoring methods, and addressing the disconnect and lack of studies between industries and academicians in the domain.
computer science, theory & methods