How can financial institutions more efficiently prevent credit-card fraud and AML using AI and machine learning technologies?
Forsström, Henrik (2021)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2021121726745
https://urn.fi/URN:NBN:fi:amk-2021121726745
Tiivistelmä
As society and payments are getting increasingly digitized and most of financial transactions are done by credit-card or wire-transfer, the focus on financial crime, fraud and Anti-Money Laundering is becoming a focus point for all financial institutions.
The aim of this thesis is to explore the methods that financial institutions can use to improve the detection and prevention of fraud and money laundering, focusing on machine learning and AI technologies, and also the human factor behind the issues that fraud detection and AML can face.
Based on the review of literature, journals and news articles, the algorithms used are quite standard and most of the innovations in that field come from combining algorithms and finding uses for these in practice.
I focused in finding experts in fraud & AML and machine learning technologies. I interviewed three experts of the subject matter to get a clear view of how to improve the prevention of financial crime. One of the interviewees was from a financial institution with AML background and the two others had technical expertise in machine learning.
The core results I found in the study are data quality and the attitude towards AML. The base data of any datasets and transactions need to be of high quality if any meaningful machine learning algorithms can be used effectively. Also the attitude of the management of financial institutions need to be improved as they usually see AML and fraud prevention as a cost that does not require investments or improvements.
All of these require large investments of recourses to update systems, train professionals and develop processes for getting this to work. Financial institutions need to change their attitude towards AML and fraud prevention in seeing it as an expense, but seeing is as a tool to prevent even higher expenses in the future.
The aim of this thesis is to explore the methods that financial institutions can use to improve the detection and prevention of fraud and money laundering, focusing on machine learning and AI technologies, and also the human factor behind the issues that fraud detection and AML can face.
Based on the review of literature, journals and news articles, the algorithms used are quite standard and most of the innovations in that field come from combining algorithms and finding uses for these in practice.
I focused in finding experts in fraud & AML and machine learning technologies. I interviewed three experts of the subject matter to get a clear view of how to improve the prevention of financial crime. One of the interviewees was from a financial institution with AML background and the two others had technical expertise in machine learning.
The core results I found in the study are data quality and the attitude towards AML. The base data of any datasets and transactions need to be of high quality if any meaningful machine learning algorithms can be used effectively. Also the attitude of the management of financial institutions need to be improved as they usually see AML and fraud prevention as a cost that does not require investments or improvements.
All of these require large investments of recourses to update systems, train professionals and develop processes for getting this to work. Financial institutions need to change their attitude towards AML and fraud prevention in seeing it as an expense, but seeing is as a tool to prevent even higher expenses in the future.