How the updated Principles of using Cifas’ National Fraud Database help its members

TruNarrative works with many global partners in the fight against fraud, but Cifas stands alone in its principled approach and not-for-profit structure. We are proud of our partnership with Cifas and see the data it provides as a key component of any successful layered fraud prevention strategy.

Cifas has now strengthened its position in the UK fraud prevention market by broadening the opportunity to use its data. As a Direct Agency, TruNarrative have built an integration unencumbered by legacy technology constraints, meaning the opportunities for the market to use this data are much greater and now accelerated by this announcement.

Our new ‘have it now’ digital society leads to completely different risk profiles, where name and address matching fraud prevention tools just don’t cut it. Therefore, a need to ‘open up’ services to incorporate new technologies and attributes is crucial.

As an industry we have a moral responsibility to share this type of information with law enforcement and with our industry peers via organisations such as Cifas.

New opportunities to fight fraud together

To use the National Fraud Database, a Cifas member must operate within the terms of the database’s Handbook – a guide that sets out eight Principles of use with accompanying guidance. These Principles and guidance describe the controls in place to protect the data on the database and ensure that the highest possible level of fairness and transparency are observed.

Cifas’ recent updates to the Handbook shows the organisation’s willingness to be more open and support us all in adapting to new risks and opportunities. There are now several possible new ways to utilise Cifas’ data, including the added functionality of using it in scorecards and predictive models. 

One opportunity to consider would be utilising the data for machine learning, but with caution. Development of state-of-the-art white box (where we understand how the models arrive at the outcome) machine learning algorithms allows users to identify existing patterns and behavioural trends that impact decision making within financial crime. Here various industry sectors can utilise the historical, multivariate patterns that exist within Cifas, constructing a profile to understand what a high-risk application would look like. In contrast, when using black box (where we do not understand the reasoning behind the outcome) machine learning algorithms, such a profile would result in no knowledge gain, as well as potentially negatively impacting the Principles.

A further opportunity lies in instances where businesses can quickly identify and counteract emerging patterns and trends by monitoring how the profile adapts over time, and how it reacts to external influences – such as new legislation. This can place industry in a unique position to predict new fraud patterns before they even occur.

These changes give the data from Cifas a new dimension, which could enable it to be used to make better informed decisions on fraud strategies. Now it’s up to us in the industry to understand how best to safely use this data in a more automated fashion, whilst also ensuring all the Principles are maintained.

The other big update is the flexibility to automatically decline – based on intelligence from the National Fraud Database. The ability to make automated decisions allows members to immediately decline facilities to subjects who are deemed an immediate fraud risk. This is another massive step forward in utilising the data from Cifas, freeing up resource from reviewing referrals that have already been deemed a high fraud risk.

We should consider that when a case is decisioned in an automated process – how can we demonstrate whether that decision would be replicated if we applied a manual process?

Combining these two principal changes to Cifas data, allows you to create a fraud strategy that is proficient and accurate. Historically combining a more manually based Cifas strategy with automated processes was challenging. But now that Cifas data can be incorporated into these automated strategies, its power for determining fraud risk becomes even greater.

An essential tool in the fight against fraud

In our opinion, the update to how Cifas members can use its data helps us be more innovative with that data, and helps Cifas’ National Fraud Database remain as one of the pivotal tools required for a robust fraud strategy. It is refreshing to see a well-established organisation embrace innovation by opening up their vast dataset to help combat the fight against fraud and cybercrime.

We highlighted Cifas as one of our key UK partners very early in our own journey, and are proud to continue to work with them. Our direct integration to Cifas means we can facilitate and support Cifas and its members in the innovative changes it has made to how we can use its data.