The bank of the future: how artificial intelligence is shaping the industry

Artificial intelligence (AI) technology can already significantly improve customer dialog, service processes and knowledge domains in the banking ecosystem. This results in opportunities such as sustainable differentiation, competitive advantages and even new business areas.

Everyone has an individual definition of what intelligence means to them personally. Overall, the use of the term artificial intelligence leads to very high and demanding expectations regarding the capabilities of a machine, but also to a multitude of fears. Can and will machines replace us in the future?

WHAT DOES COGNITION MEAN?

The concept of human cognition, derived from psychology, is illustrated by abilities such as creativity, memory, imagination, will, humor or the ability to learn. When we talk about cognitive systems today, we claim to be able to represent one or more of these cognitive abilities with IT. This leads to the definition that a cognitive system possesses "behavior-controlling human abilities". Compared to the term "artificial intelligence", the term "cognitive system" therefore describes a machine that can reproduce individual human abilities much better.

One particularly relevant cognitive ability is the ability to understand and interact with natural human language. The view is very often held that artificial intelligence only exists if you can have a "human" dialog with the machine. However, as natural language is highly complex and unique in its individual characteristics, this has been difficult to implement by machine to date. In natural language, it is important to take into account the levels of meaning (word meaning, sentence meaning, context of a statement) as well as the context and possible ambiguity (ambiguity of words, example: I can be sitting on a bench or I can be in a bench).

DATA AND COMPUTING POWER: WHY IS KI SO RELEVANT TODAY?

In recent years, the growing power of computers and the wealth of available data have ushered in a new era of systems with cognitive capabilities. Intensive analyses and complex analytical methods are now used to develop algorithms that can learn and continuously improve with the help of additional data. The term "machine learning" has now become firmly established.

Technologies with cognitive capabilities are already often an integral part of our lives: we interact with assistants such as Siri or GoogleNow on our smartphones throughout the day; when shopping online at Zalando or Amazon during our lunch break, we are offered products tailored to our tastes; on the way home, we listen to automatically compiled playlists from Spotify and in the evening we follow our individualized entertainment offerings on Netflix.

Despite all these advances, artificial intelligence is far from replacing human capabilities in many situations in life today.

WHAT IMPACT DOES ARTIFICIAL INTELLIGENCE HAVE ON BANKS?

Digitalization has not spared the banking industry and has led to some radical changes. You only have to look at the number of FinTech companies in Germany to know that the industry is currently undergoing extreme change. In the course of this, I would particularly like to highlight the interaction between banks and their customers, which has changed dramatically as a result of digitalization. Customers now expect a different and, above all, better, more personalized customer experience (user experience) than they might have done a few years ago.

In addition to the challenges posed by digitalization, there are also many opportunities for the banking industry. In addition to the user experience already mentioned, banks can use artificial intelligence in particular to improve and personalize interaction with their customers, but also to improve efficiency and thus their own cost structure.

In customer advisory services, cognitive systems are used in direct dialog, but also to support advisors. This is known as cognitive banking, and virtual advisor and chatbot solutions can be used to provide both simple and more complex advisory services. In a simple application scenario, for example, a machine chat dialog can be used for simple service requests. In more complex scenarios, a virtual advisor can provide advice on Construction Financing or pension protection.

However, cognitive technology does not necessarily have to be used in direct dialog with the customer, but can also be made available to the bank advisor as an internal tool. As a concrete example, a cognitive dashboard can automatically provide the bank advisor with relevant advisory information before or in real time during the consultation or even make advisory recommendations. In this way, cognitive virtual advisor solutions can significantly reduce preparation times and research for customer appointments and ensure a targeted and, in some cases, higher quality of advice.

Cognitive technology is already being used today to optimize service processes in service centers. By understanding natural language, service processes can be (partially) automated. This includes both the routing of recognized customer concerns to the right process function and support in finding the right problem solution. As soon as a customer issue is automatically understood and processed fully automatically, we speak of cognitive robotics solutions.

SELF-LEARNING SYSTEMS AND THE LIMITS OF KI TECHNOLOGY

Artificial intelligence is subject to high expectations, which often cannot yet be realized in all dimensions during initial practical testing. It is therefore important for the use of AI technology in a banking environment to understand how a system that uses artificial intelligence works and the sequence in which cognitive technology should be introduced.

Analogous to a person's learning process for individual human cognitive skills, cognitive technology consists of individual disciplines (e.g. linguistic understanding, analytical skills, understanding of tonality and visual representation).

Building on this, a cognitive system is first trained with banking expertise and prepared for autonomous "further learning".

A self-learning system must "perceive" the world, collect data, understand the collected information and act in order to make well-founded recommendations. To do this, it is necessary to collect the right amount of relevant information, process it and "train" the system in several steps. This process may still seem comparatively cumbersome today, but with increasing data examples, an AI system can and will learn independently and continuously improve itself.

THE HUMAN COMPONENT WILL REMAIN IN THE FUTURE

In the coming years, artificial intelligence will continue to revolutionize digital customer engagement through natural language, research questions and issues that cannot yet be answered, facilitate decisions through evidence-based recommendations, and automatically check and analyse legal compliance and regulatory requirements.

Systems with cognitive capabilities are changing the role of the employee in a bank today and in the future, but will not do away with the human component in the bank. Human intelligence is and will remain unique in the coming years.