In order to be able to answer this question, we should take a look at how banking has evolved over the years and how it is being conducted today. I will go briefly over what AI is and also talk about the drivers and barriers of AI in the banking industry.
Let us start by stating what AI is. AI is software used to address specific problems for the purpose of making smart and automated decisions and predictions to assist us in simple and complex tasks.
AI is not traditional automation. With automation, we simply use predefined functions to run specific tasks at high speeds with all sorts of validations in order to eliminate human errors and generate the needed outputs.
As with AI, we want the software to provide us with answers that we do not anticipate based on human-like reasoning utilizing huge amounts of data and complex correlations that are practically impossible for a human being to analyze.
We experience AI on a daily basis without paying attention to it. We experience it with our junk mail filter, our interaction with social media (such as Instagram determining which posts to show us based on our past behavior and preference), how we have access to our phones through face recognition, chatbots, virtual assistants, etc. not to mention self-driving cars and factory robots. I do not think we can imagine our phones without such login security any longer, and that is exactly what is going to happen to us every time we are served through AI in the future.
No one will accept to wait on the line to receive any type of service. Eventually, everyone will be expecting to be served immediately around the clock once a question comes up or a need arises.
But what are the core technologies of AI? They are Machine Learning, Deep Learning, Computer Vision, Natural Language Processing, and Forecasting and Optimization. Computer vision, for instance, is the ability of a computer to recognize an object. Natural Language Processing is the ability to understand speech.
It is not our objective today to dive deep into AI and explain how Machine Learning or Deep Learning works, but it sure is important to at least have an idea about them. With Machine Learning, software will be able to detect certain patterns in data using predefined algorithms and through training. For instance, the machine (software) will be able to detect a fraudulent activity by simply analyzing data, even if this type of fraud was never defined before! Deep learning is simply a more complex type of machine learning which uses the neural network approach.
On the other hand and in order for AI to work, we need several underlying technologies to carry on the processing including access to large storage for processing data, cloud computing, edge computing, and open source technologies.
But why do banks need AI? The answer is very simple: Banks need AI in order to provide a better customer experience! They need to be so responsive in order to meet their clients’ expectations and by applying AI, security needs to be drastically improved to give clients the assurance they need to rely on machines.
Are these enough reasons to force banks to dive into the AI realm? The answer is YES. Banks need to provide an enjoyable customer experience if they are to keep their clients.
Banks’ biggest competitors are not other banks anymore; they are technology companies. For instance, Facebook, Google, Amazon, and other big companies are now offering banking services to their clients. Banks need to compete against these companies and start offering better online services to keep their clients. Clients expect immediate services; they don’t want to wait and want immediate answers. They do not want their instructions to take time to be executed. All of this require a different type of automation.
For the past decade, Banking services have become more online and accessible through our mobile devices. However, online and mobile banking no longer offer an edge. Clients need to have a virtual assistants (AI) along with online and mobile banking.
Banks need to provide advance fraud detection mechanisms in order to secure clients’ finances and need to provide financial advisory to their clients using data analytics (AI) for better results. All of this require AI.
But the road to AI is not easy. There are so many barriers that make it very tough to achieve. Among all the listed barriers, I would like to stress on the following:
1. the lack of resources since AI is still rather new and expertise is scarce,
2. the risks associated with AI decisions such as the lack of explainability,
3. access to data,
4. missing legal framework, etc.
However, all of these barriers will be overcome in time and banking will not be the same.
Banks will have to set a strategy, plan for it, develop their ecosystem and go through a cycle of development, implementation, monitoring, and corrective action.
As mentioned before, Banks have already started using AI. It is mostly used for risk management, enhancing customer experience, and quality control. As we speak, banks have the following services based on AI:
I take as an example to explain how the process works as AML is an important application of AI. Traditional methods utilized rule-based engines to detect ML. With AI and machine learning, AML detection results have greatly improved. EXPLAIN GRAPH.
We can take a look at the different activities at a typical bank and see how much AI has been implemented for each sub activity. We can see clearly that AI is still in the early stages since the activities that rely most on AI are still around 60% at most relying on AI technology. We still have a lot to do and learn before AI takes over.
Talk about investment (from less than 10B in 2016 to almost 100B in 2025 expected to each 300B in 2030!)
In the future, we should expect to go more virtual. Traditional banks will almost disappear. Digital and virtual banks will take over. Payments will be frictionless based on wearables and biometrics. Digital currencies will be adopted and AI will dominate. People will start trusting machines more when it comes to their finances exactly like we trust a calculator when we multiply 2 large numbers. In a couple of decades, AI will change the way we live, and banking is only a part of this new way of life.