How banking is using artificial intelligence to pick up financial distress

Phil Grady, chief executive at Castlight Financial, one of a number of fintechs working to reduce financial distress. Picture: Contributed

And there are few fields where AI is driving innovation as quickly as in the thriving fintech sector, an area in which Scotland is rapidly building a world class reputation.

Financial institutions of all sizes are developing and implementing advanced AI systems designed to pick up and prevent financial distress among customers.

This includes the likes of Royal Bank of Scotland, which recently opened a research unit dedicated to solving real-world problems with data-driven innovation.

In its broadest definition, AI has been with us since the advent of machine learning in the 1950s. But today’s tech, based around systems capable of not just holding data but understanding and interpreting it intelligently, opens up a world of possibilities.

There are two main reasons why AI is now coming to the fore, explains Stephen Ingledew of FinTech Scotland: “The first is the power of technology to manage a huge amount of data and understand it. The actual machine becomes intelligent.

“The second reason is that there’s far more data. We’re all leaving a digital and data footprint everywhere we go, every time we do something.”

This combination has allowed firms innovating in the banking space to build AI which can flag vulnerable customers or those at risk of potentially heading into financial difficulty.

Some banks, using voice recordings from historical cases where the customer has gone onto suffer some form of financial distress, are training algorithms to spot tell-tale signs among current customers. These individuals can then be offered additional support to proactively address any potential problems.

This is also beginning to take place at a commercial level, tracking patterns in businesses which later fall into administration.

“AI can support horizon scanning to identify early warning signals for commercial and corporate customers using a combination of transaction and news data, and predictive algorithms,” says Chris Brown, senior consultant at Deloitte.

One of the most prominent Scottish fintechs working in this space is Castlight Financial. The Glasgow firm, acquired by Experian last year, uses data and AI to support its “Affordability Passport”.

This digital tool helps financial retailers loan responsibly by allowing their customers to share a real-time picture of their financial situation, including income, expenditure and financial capability. It gives a comprehensive overview to ensure customers can truly afford products, with the aim of cutting off the potential for financial distress before it begins.

Castlight's tech can also allow customers who may otherwise be “locked out” of the credit market, perhaps due to having a poor credit score if they are young, or have returned from working abroad, to demonstrate what they can afford and help them access loans when needed.

On the commercial side, business to business commerce specialist Previse, which has operations in Glasgow and Edinburgh, has developed an accounting tool which enables major companies to pay suppliers instantly on receipt of an invoice.

It aims to tackle the much maligned issues of late payments which afflicts the vast majority of small businesses. Previse’s AI-powered analytics read invoice data to recognise and flag “problematic” invoices, alerting the business to potential red flags and allowing the rest to be paid instantly.

As the financial sector relies increasingly on AI to tackle social issues, building and maintaining customer trust in AI services will be key to growth for financial institutions, says Brown.

He said: “AI is a fundamental building block for a large number of fintech companies in Scotland. As it becomes more mainstream, through direct-to-consumer products such as smart speakers or our growing mobile phone capabilities, it’s crucial that companies can provide a level of transparency to develop trust with customers on how AI is being implemented.”