Harnessing alternative data in the fight against fraud

By Carol Hamilton, Director of Compliance & Fraud, EMEA, GBG

28 September 2020

The recent global crisis has set off a major fraud resurgence. With the world continuing its acceleration towards becoming digital-first, and with everything from work and transactions to entertainment and shopping happening online, potential attack vectors and opportunities are exponentially growing. The UK alone has seen a 66 percent rise in scams during the pandemic so far.

This is especially true for the financial services sector, as banks and financial organisations quickly shift their operations online during the pandemic in order to reach newly remote customers. Actions such as onboarding and sensitive transactions have been forced to take place purely remotely, while ID verification methods had to be adapted to cater to remote customers – during lockdown, the FCA even announced plans to accept selfies as part of a holistic identity verification process.

Fighting fraud with traditional techniques is no longer enough. As fraud becomes digital-first, so should anti-fraud techniques - businesses need to combine technology and data to create intelligent, real-time responses to problems, without a customer, or potential fraudster, ever even knowing. To do this, alternative data and machine learning are quickly becoming go-to solutions.

How can alternative data combat fraud?
It’s no secret that data holds crucial value in today’s business and financial services sectors. When it comes to fraud, leveraging data has the potential to help teams spot unusual activity, flag risks and predict fraudster activity.  

And as the fraud landscape grows in complexity, we are also seeing new alternative data being leveraged across fraud and compliance teams. Alternative data goes beyond traditional data sources to encompass new areas such as live locations, consumer behaviours and social media. For example, in financial institutions, to make fast and accurate decisions about fraud risk, behavioural data drives an incredible amount of value – data points such as smartphone device attributes, travel activity and typical website use pattern etc.

With alternative data, fraud prevention teams can look at various factors to identify unusual, potentially fraudulent activity. Has a sim card recently swapped phones, indicating false activity? Does the ‘customer’ typically pay for purchases on time and what are the usual spending patterns? Does the transaction location match the customer location, or does the end destination appear unconnected to their usual activities? Are there only three contacts on their device – could it be a burner phone?

Frequently updated, accurate alternative data sources can ultimately help to understand daily activities of an individual and create an overall view of their behavioural patterns. It’s only by considering multiple layers of data that teams can create a cohesive, data-driven risk alerting process and support accurate end decisions.

Combining data with emerging technologies
As data becomes a central component to fighting fraud, the introduction of Artificial Intelligence (AI) and Machine Learning (ML) will be crucial to driving successful results. In the financial services sector in particular, there are hundreds of thousands of transactions taking place per second. AI and ML are therefore helping teams to manage and understand the mounting data sets, and use them effectively to spot undiscovered patterns, which is crucial for flagging potentially suspicious behaviour.

It's important that all AI/ML technologies are integrated within the investigation environment – rather than working in silo. In doing so, assessed patterns of behaviour can be automatically fed back to help machines adapt, automate and train themselves.

Teams must also identify powerful, suitable algorithms and proven techniques when using AI/ML if they’re to add value in the fraud space. Even in the most challenging fraud environments, the right algorithms will be the stepping stone between success and failure.

A multi-layered approach is business-critical
At the end of the day, it’s impossible to base fraud decisioning on singular factors. Teams need to gather a comprehensive view of potential threats and understand the context surrounding their decisions.

Take a simple bank account application - data analysis may indicate there are a number of low-medium risk flags that, when combined, all roll-up to a ‘red’ flag which may warrant an investigation. For example, if the age of the applicant is 18, but the salary is stated as £300,000, that’s a potential anomaly. If their address is a fraud hotspot, their IP address is connected to an unusually high number of applications and their phone number has recently had its sim swapped - which indicates a burner phone in use - a level of suspicion comes into play. It’s only through combining new technologies and data solutions to generate these all-encompassing views, that teams can work to accurately detect fraud attempts. 

As well as utilising these new, alternative data sets and advanced technologies, we need to educate consumers on the dangers and warning signs of fraud. At the end of the day, preventing fraud will be more successful if the process begins with the end-user. We should therefore strive to combine advanced fraud prevention technologies with improved consumer awareness in the fight against fraudsters.

Become a bobsguide member to access the following

1. Unrestricted access to bobsguide
2. Send a proposal request
3. Insights delivered daily to your inbox
4. Career development