The dual role of data science in fighting financial crime

Media release

3rd November, 2020, posted in Global Banking & Finance Review

PwC’s Global Economic Crime and Fraud Survey 2020, which surveyed over 5,000 respondents over a 24-month period, found that an average of six frauds per company was reported, costing a massive $42 billion. It also found that just 56% of companies undertook an investigation into their worst fraud incident. The message is clear – fraud is a very real threat, it’s on the rise, and to echo the view expressed in the PwC survey report: “It’s a risk you ignore or underestimate at your peril.”

This kind of financial crime is a worrying trend, and is a key reason why fraud prevention services are looking to data science for help. Working with fraud prevention services, data science can help build and enhance existing intelligence capabilities, as well as further deepen understanding of the crime networks involved in fraud.

When it comes to fighting financial crime, data science has a role to play in two main areas: the technology and tools used to identify fraud, and the education and training necessary for the teams to use these tools to maximum advantage.

  1. The (Data Science) Tools of the Trade

Data science uses advanced analysis to deliver value from data by enabling informed decision-making. To illustrate this in a financial context, a data science initiative may include reducing the number of false positives to improve matching when searching the National Fraud database, improving fraud prevention services’ members’ understanding of the data, and providing further intelligence based on data – such as emerging crime patterns.

By using data science capabilities and expertise in this way, fraud prevention services will be able to utilise this intelligence more effectively and efficiently to help identify current and emerging fraud threats. This data will then be used to inform members and the wider fraud prevention community.

Another way data science can be used to combat fraud in insurance, for example, is through identifying fraudulent claims. With regard to these, data science can help detect claims that look “unusual”. Artificial Intelligence (AI) techniques are used to examine all elements of the claim in real time and match it up with the history of similar types of claims that have been made before. If there is an anomaly, the claim will get pushed to one side for further investigation. This means that insurance companies can ensure they are still able to pay claims quickly, whilst checking thoroughly for fraudulent claims.

  1. (Data Science) Knowledge is Power

From a data science perspective, having the right skills are essential for companies tackling financial crime in order to identify the increasingly sophisticated digital crime frauds and scams being perpetuated across the industry. Financial services fraud teams have been striving to detect fraud for a long time, but not necessarily using the latest, data-driven techniques – many are still immature in terms of their data capability.

In a time of huge advances in technology, new capabilities around the interpretation of data and an enhanced acceleration towards digital transformation partly as a result of a global pandemic, data science offers the strongest solution for companies fighting financial crime. Traditional approaches are no longer enough, and it is essential that financial crime teams build data science skills through education and training if they are to keep up with the growing tide of cyber-crime.

When it comes to fighting financial crime, data science is the key to tackling an evolving financial landscape, governed by complex compliance requirements, to prevent increasingly sophisticated criminal attacks.

By Tim Oldfield, Financial Services Client Director at Mango Solutions