How will AI and Machine Learning change the game for the banking sector in 2020?

By Stephan Steyn - Solution Architect for Hybrid IT at Hewlett Packard Enterprise South Africa

Data analytics is nothing new. We’ve been doing it for decades to figure out why things happen, when they will happen and what we can do to influence the way things happen in the future. What is new is the number of different sources we have, the sheer volume of data, and the speed at which we are able to analyse this data to deliver valuable business insights. 

That’s where the concept of deep analytics comes in. We use highly sophisticated mathematical algorithms when analysing huge pools of data to find connections between and within the data. These algorithms harness Machine Learning (ML), so that the more data we provide, the smarter the analytics and the better insight derived. Artificial Intelligence (AI) and ML have the potential to revolutionise so many areas of business, including the banking sector. 

The trouble with an algorithm that learns from itself is that you have to be extremely careful what data you give it to learn from. Microsoft’s Tay AI bot, for example, was an experiment using AI to learn from the conversations of users on Twitter. Within a day it had become a racist, sexist bigot and the experiment had to be shut down. 

You also need to be extremely careful to clearly define the problem you want to solve. The supercomputer, Deep Thought, in “The Hitchhiker’s Guide to the Galaxy” is an excellent analogy to demonstrate this. In the story, advanced galactic ‘beings’ build a gigantic supercomputer and ask it to find the answer to life, the universe and everything. After many years of analysis, it comes up with the answer ‘42’. This highlights the need to fully understand your business problem before you apply any sort of deep analytics to it. 

These are two reasons why we haven’t yet seen much in terms of AI and ML in the banking space. It’s not because they don’t see the potential. The opposite is true, and the larger banks have all invested heavily into the technology and have been working furiously in the background for years now to develop and implement their solutions. 

One area where the benefits are obvious is in the prevention of financial crime, but this isn’t something the public will see as it’s all internally focused. For end customers, AI and ML will come to life in the form of improved customer experiences. This is a significant point of competitive differentiation for these banks. 

Some aspects of AI have already been revealed, like the WhatsApp chatbots implemented by both Nedbank and Absa. These bots help users to have a ‘human-like’ experience when finding information, without the need to actually employ people to give the service. We have also seen ‘voice password’ systems implemented that use voice recognition to verify a customer without them having to run through security questions. Deep learning algorithms can also be used for per-customer personalisation on marketing of new products and services, or on the products and services themselves. 

We could also see AI-enabled cameras at ATMs with video analytics to detect potential dangers such as theft and heists. Banks all already have pictures of their clients on file, so they could easily run a facial recognition tool to ensure that the person drawing the money is the account holder. This could even go so far as to detect emotion and understand whether someone is under duress such as being forced to draw the money against their will, such as if they are being held up. 

The Internet of Things (IoT) opens up a whole new world of possibilities, since banks already have connected devices everywhere. The next step is to bring computing closer to the edge to deliver the power needed for video analytics and real-time results. 

One thing is for sure, 2020 is going to be an interesting year. Many of the projects that banks have been working on for a number of years will come to fruition, and a lot of the focus will be around improving the customer experience. Exactly what we will see, only time will tell, but the fact is that AI and machine learning are game changers in the banking sector.