Boris Huard, EMEA MD for Identity & Fraud at GBG, the digital identity experts

Fraud is a huge problem with fraud soaring to its highest ever recorded levels, and the pandemic further exacerbating the onslaught due to greater online dependency. Scams, phishing, account takeover, third-party identity theft, and other types of cyber schemes accounted for the highest fraud losses in 2020 amongst the 900 financial institutions we surveyed*.

Against this backdrop, artificial intelligence (AI) and machine learning (ML) are often seen as the silver bullets in the fight against fraud, but for smaller and fast-growing financial institutions, the solution is actually much closer to home and faster to implement.

The AI implementation gap

There is strong intent amongst financial institutions, including banks to use AI to reduce fraud. In our own independent research, more than half of financial institutions plan to use AI to reduce fraud in 2021 and beyond, and almost 60% said they see AL and ML it as key to being able to spot new patterns as they evolve.

And that is key – fraud is evolving all the time, increasing in pace, traits, and sophistication. Staying one step in front is a challenge and while the use of AI is enabling large organisations such as Visa** to prevent billions of pounds worth of fraud every year, the rate of adoption and maturity in the use of the AI varies widely across the industry –

mostly driven by the size of the organisation and nature of services.

AI as a distraction

The reality for many smaller, new and fast-growing financial organisations is that they don’t have the in-house resources needed and the time to introduce and fully benefit from complex AI and ML at this present time. Instead, they need to look at what they can do now as well as in the future. 93% of financial institutions agree that managing fraud losses is a key necessity and focus for business strategy. To reduce fraud losses right now, these organisations must take stock of the data they already have, while introducing new data sets, so that they can get a clearer understanding of their customers – ultimately ensuring they can let the good people in and keep the bad people out.

With better access and use of additional data sources such as mobile phone data and IP addresses, behavioural data, real gains can be made in identifying and combatting fraud.

Better data = faster action

Speed of decisioning is often the crucial element in the growth of more digital-first institutions. The ability to make fast decisions without negatively impacting the customer journey is fundamental to this. Smaller and more agile firms have speed on their side to bring in new data, with greater flexibility to test existing and new data whilst adjusting fraud detection and identity verification rules. An advantage that larger organisations and those with legacy systems often don’t have.

Tips for becoming more data-focused in the fight against fraud as you grow

Give fraud a seat at the top table

Fraud isn’t just an issue for fraud professionals within your organisation. Fraud will seriously dampen and derail any business growth ambitions, and has long-term financial and reputational ramifications. Smaller and new organisations, in particular, are especially vulnerable to reputational risk and having the right anti-fraud mindset will be essential as an organisation moves towards being regulated. Fraud prevention needs a seat at the table but it is your data that needs to speak first.

Understand your data

Start by identifying the scale of the problem and the challenge – does your current data meet your needs for robust identity verification and fraud prevention? Where are the gaps and what other sources of data need to be within your fraud toolkit so you can better spot, at speed, potential fraudsters, and fraudulent activity. Can you analyse the data that you already have?

By better understanding the two or three critical data sets that impact your fraud detection levels, you are in a much stronger position to undertake the right analysis of the right data before building on this to identify, join up and plug in extra data into the mix. Adding these extra layers of data will bring out more predictive intelligence to act upon.

Build trust in a digital world

Consumers are looking to financial institutions to take a lead on fraud and to better protect them. There is an opportunity to further differentiate yourself beyond product and good marketing – build consumer trust and be transparent in the fight against fraud. It is the consumer that pays the price for fraud as typically fraud losses are passed on by way of the rising cost of goods and services. Can fraud savings be shared with your customers?

Put yourself in the shoes of the consumer and the financial and emotional implications of fraud – what data will help them to help you? Consumers are increasingly open to wider use of data and biometrics if it means they can be identified and verified much more swiftly, and are better protected against fraud. Good fraud prevention and use of data builds trust.

Consortium and collaboration

Don’t underestimate the importance of data sharing. Whilst there are a limited number of fraud consortiums or networks, the bigger banks and insurers have been data sharing for many years. Whilst other smaller financial institutions, don’t have this advantage, there are lots of opportunities to foster and benefit from greater data sharing with suppliers, customers and competitors. Learn from the banks and insurers, and look to other sectors to see how they use data to identify and reduce fraud.

The consortium approach is extremely powerful and effective when it is applied to cross-border fraud. Fraud has no respect for geography and with the bulk of fraud taking place online, working in greater collaboration and sharing data means you are more likely to be able to spot trends and repeat offenders who are organised across borders targeting smaller financial amounts but at increasing frequency.

Test and enhance

Ultimately, if you get the data right, the rest will follow. Data needs to be at the heart of any good fraud prevention strategy and having a more data-focused approach means you can make the most of what data you already have. Take steps to bring your data together, test and enhance with the addition of alternative data sources, so that you have the value, volume and speed from your data to make more accurate fraud risk assessments – preventing fraud before it impacts your business. It is only then, when your data is firmly in place as the fuel to drive better fraud prevention decisions, that any intent towards AI and ML should become reality. They have a big part to play in the fight against fraud but only in combination with good data, never alone.

Data needs to be your priority.

*GBG fraud survey 2020: smoothing the customer journey and preventing fraud 901 fraud professionals within financial institutions surveyed

**Visa prevents around $25bn in fraud each year, using AI technologies (Visa / TechCentral.ie)