However, it is no longer enough for businesses to merely provide customers with digital experiences. 

Consumers are increasingly calling for a service that is hyper-personalised and tailored to their specific needs — similar to the experiences they already have with apps such as Spotify and Uber. As a result, new technologies are constantly emerging, some with more hype than justified. Explainable AI (XAI), however, looks set to be an exception.

The banking sector, like every other, is feeling the pressing need to move towards bespoke solutions with greater transparency — as a result, there is increased appetite to leverage the power that XAI can bring to banking. 

The use of AI in decision making has received criticism recently for its perceived biases and lack of transparency surrounding its outcomes. There has also been some scepticism among consumers, who can be reticent to share their personal data. Yet it is only in doing so that they can receive the personalised services they crave and that provide them with what they actually need. 

Fortunately, XAI models can help overcome these concerns, illustrating to customers the value in sharing their data, while also reassuring them that it will be used in an appropriate and non-biased way. 

When it comes to delivering personalisation at scale, the solution is ‘white box’ XAI systems, which help to shine a light on the AI decision-making process and offer a new level of transparency into how the system operates when mapping profiles, scoring credit ratings and making decisions that involve customers. 

While too often the decisions reached through AI and machine learning can feel remote and automated, XAI systems demonstrate how and why certain decisions are made and encourage customers, regulators and banks alike to trust in those outcomes. These systems also allow for problems to be amended as soon as they occur. 

XAI systems are also allowing banks to carve out new revenue streams. For example, with valuable insights into a particular AI outcome, banks can reroute customers that have been declined for a particular service towards a more suitable option for them, and for which they would qualify. This not only allows banks to provide highly personalised experiences, but highly effective services to their customers.

At Temenos, we have patented an XAI model that is now a default part of all our technology and any model we build. For example, if we build a credit decisioning model, our XAI technology enables us to see all the factors that contributed to that decision and allows users to interact with it.

We are only just scratching the surface of what XAI can do, but as banks look at how they can leverage its capabilities, it will increasingly become integral to product development. With regulators joining the discussion surrounding explainable and transparent AI models, 2020 is set to be a big year for XAI.