How is AI going to impact the finance sector

The role of AI is one of the dominating conversations across industries now, and there's little doubt it will continue to influence both commercial strategy and common parlance in equal measure. However, while its presence is undeniable, the best use of it in different industries remains open to debate, so it seems prudent to add our ten pence worth to the conversation. 

We've been around long enough to be curious, but never overexcited, about new innovations. We embrace progress, but retain a reverence for due diligence, and the pillars on which good business stands strong. We are also vocal as a business about our commitment to human underwriting over automated decision making when it comes to funding approval.  

Therefore, in this article we don't seek to provide answers so much as to ask questions about the role of AI and how it has the capacity to impact the finance sector as we see it.  

The positive potential of AI in finance

Perhaps one of the key things should be to look at the scope of AI application before we get into its merits. There are lots of different ways in which we all engage with AI already - not least, chatbots, virtual assistants, and AI agents.  

Chatbots are perhaps what most of us are familiar with, ranging from the fairly simple, rule-based versions that answer straightforward questions, to the AI-powered ones which can handle more detailed queries in real-time. AI virtual assistants simulate human language to deliver responses to queries and also vary in the complexity of their capabilities. These tools are often used to supplement customer service by answering FAQs, triaging customers into pathways, or enabling an element of self-service. 

Meanwhile, AI agents are amongst the more advanced end of the spectrum, interacting with the environments around them, gathering data, using it to perform certain tasks to meet particular goals, and some with quite an advanced capacity for learning. These offer arguably the greatest opportunity in terms of managing tasks efficiently and accurately and supporting business scalability.   

With all of this in mind, the positive potential of AI in finance is well documented. It seems there's lots of possibility when it comes to security, risk management, some degree of personalisation, efficiency, and scalability. That said, some of the areas in which it can offer benefits are also areas of potential exploitation, whether it's cybersecurity risks, data privacy, regulatory compliance, or bias in algorithms. There are things that can done to mitigate those risks - ultimately, as things stand it appears to come down to balance, and perhaps no small amount of common sense. 

However, a lot of the areas of potential don't really seem to be 'there' yet - they are still comparatively rudimentary - although the pace of change is no doubt rapid. In part, one might also argue that what isn't fully there yet, is the human capacity to make best use of the tools available. Either way, awareness of the limitations as well as the possibilities are worthwhile. 

Exercising caution with automation and finance

Our Non-Executive Director, Stephen Bassett, who is also the Chairperson for The Guild of Business Finance Professionals, recently wrote on LinkedIn: 

"AI is a useful tool for some things, so use it to check things if you must, but not to change things.  Remember, it has no imagination, it cannot innovate and everything it does is derivative. You may think its output looks and sounds reasonably clever, but it is sourcing other peoples' and machines' previous output and is usually easily recognisable as such." 

This seems a good place to start exercising caution on the topic of AI. Finance, certainly in the area that we operate in - providing business funding solutions - is a customer service-based industry. While there is an element of box-checking that goes on, and indeed much of that can be automated, there is also a human story - and indeed a business story - behind most applications that needs some consideration. There is no doubt that AI is incredibly clever, but as things stand it is only as clever as we make it - and crucially, it isn't human.  A good example of this came in our own sector recently, when a client who had previously been turned down for a loan with another lender, had their application approved with Rivers within 48 hours. 

The online company specialised in providing leads for the gambling industry. They came to us looking for a medium-term business loan of £60,000 over six months to invest in website development and content strategy. The company had previously gone to another lender, known for using an automatic portal for its initial decision-making. Having initially approved the loan, when the lender's team saw that the business was linked to the gambling sector, they instantly declined it.  

At Rivers we are non-sector-specific, and do not discriminate against businesses that many other lenders consider high risk. Instead, we always consider the loan application based on the strength of the company itself. The business had a strong track record, and their application had everything we needed to consider it properly. Furthermore, we do not use automated systems - our human underwriting team take the time to understand the business and the reason for the borrowing, doing all the necessary checks and due diligence to ensure we are lending responsibly.  

Having submitted the application on a Tuesday, the broker asked if we might be able to release funds by the Friday. Thanks to the efficiency of the broker and the company in supplying everything we needed, as well as the strength of the application, we approved it and had funds in their account by the Thursday.  

Artificial intelligence vs human intelligence

In the AI debate, artificial intelligence and human intelligence are often pitted against one another. To us, that seems to miss the point - surely, they are different, and they serve different purposes?  

Take the conversation away from finance for a moment. When a colleague wanted to visit a restaurant for a series of meetings recently, he contacted his chosen location via their website to see whether he needed to book a table or simply turn up. The chatbot greeted him with deceptively human grace and said he needed to go to a sister website to book a table. On the sister website, another bot told him he couldn't book a table. Having been sent around the houses, he eventually phoned. The chatbot had not been able to handle a question beyond very simple parameters, fundamentally undermining the customer experience and leading to irritation.  

Naturally, not all AI is created equal, and it only stands to get more intelligent, but we can only exist where we are today, and as it stands, AI has its limitations. Those limitations are not just about the AI itself, but human interaction with AI. The capacity for misinterpretation and misapplication, especially in critical sectors such as finance, medical advice, legal guidance, news, and safety, is great. 

That doesn't mean that we shouldn't use AI, but to our minds, and in our business practice, it does mean that we should be wary of relying on it, and certainly relying on it unchecked. There is a hypothesis that one shouldn't see AI as a tool, but as a colleague. If we are to take that approach, then there are very few contexts in which we would only have one set of eyes cast on important pieces of work produced by a colleague - so it stands to reason we wouldn't treat AI differently. 

At Rivers, we like to be forward thinking whilst recognising that our customers are not computers and don't always fit into a perfectly automated box. Our decisions are ultimately made by people, so when you receive an approval from us, you know it's something you can depend on.  

If you would like to find out more about Rivers’ finance solutions, from asset finance to business loans, you can speak to the team any time. 

Contact Rivers Finance Group