Exclusive interview with IBMs Vivek Bajaj on how AI is used to disrupt financial services

As the world of money is constantly changing, technologies such as Artificial intelligence and the Internet of Things have the potential to enable businesses to extract more value out of data and utilise it to provide a better service for customers. Moreover, leveraging these technologies and combining them with human expertise have resulted in the creation of new financial ecosystems.

Could you share with our readers what IBM Watson is, the idea behind the system and the technology that it uses? Are there any other similar tools/initiatives on the market? If so, what makes Watson stand out?

IBM Watson is a unique AI platform developed in 2011. It is a question-answering computer system capable of answering questions posed in natural language, named after IBM’s first CEO, Thomas John Watson. The computer system was initially created to answer questions on the quiz show Jeopardy and, in 2011, the Watson computer system competed on the show against legendary champions Brad Rutter and Ken Jennings.

During the show, Watson interacted directly with humans, by receiving natural language questions and delivering an instant response, quicker than the two champions. It was very good during the initial demonstration, however technology needs to be put to work in real world. As a result, it was firstly applied to health care, for the treatment of cancer. Currently, Watson is actively used in multiple cancer research hospitals, and operates as a doctor’s assistant.

When it comes to financial services, IBM Watson can be applied to specific areas such as financial crime prevention, regulatory compliance, and payments. What makes it so unique is the deep amount of research and development expertise we put in there and the application to specific business problems, as opposed to a general AI system, which can be used in multiple different contexts. A critical element of AI systems is the data on which they are trained – it’s that combination of innovative AI capabilities and deep domain expertise – be it medical training and experience or regulatory expertise. This one is very specific to solving different business problems.

How is IBM, with the help of AI, transforming the financial services market?

As I mentioned before, KYC, AML, and payments are some key areas where Artificial Intelligence is used. Despite its acknowledged benefits, this technology needs to be combined with real domain expertise. For example, IBM is working with a number of banks, such as Mizuho Bank in Japan, to improve KYC. In addition, IBM acquired Promontory Financial Group in 2016, who bring unparalleled expertise in the area of regulatory compliance and financial crimes.

At the moment, according to a number of examples from clients we worked with, businesses spend approximately 15 minutes to onboard a new customer. The current KYC process is using data that the business has purchased from different sources, and combines it with some primary research performed on Google on different sources. However, by using AI, you are first and foremost enabled to look at relevant news, for a specific person and then automate that creation; secondly, AI doesn’t remove the human from the KYC process. Instead it is doing the preparation work, and then the human being is reviewing the content, and approving the KYC record afterwards. So even the regulators understand that as long as you are using AI as an assistant, it is more effective than replacing the human completely.

From the AML perspective, whatever AML system you are using - and there are many in the market whether it is SAS, Norkom, etc. or any of these systems - from my experience in working with major banks, they are generating an enormous amount of false positives. Almost 90% of fraud alerts are actually false positives; however, well-trained AI systems can better understand relationships and behaviours across extended data points and interactions to help detect some patterns in these false positives, and reduce them. And, in addition, improve the detection of true positives.

Can predictive and cognitive analytics be woven into customer interactions, to personalize and enhance user experience, and if so, how?

There is a lot of hype around customer interaction and service. Probably many of us may have seen or heard the talk the Google CEO did recently on stage when an AI system was used to make an appointment at a hair salon or at a restaurant. It is fine, however there are some key points to keep in mind in financial services.

Firstly, it is important to not misrepresent things, it is very important to ethically make sure that people understand when they are interacting with an AI assistant and when they are interacting with an actual individual. And most people are fine with that as long as they actually know it.

Secondly, in terms of using customer interaction bots, it is not only about Amazon-like type of interaction with Alexa to order products, it is really about deep domain-specific understanding. Therefore, we enable businesses to apply AI bots internally in specific areas, whether it is wealth management or retail banking, to improve expertise and interact in a customer service bot way.

For instance, IBM is working with Bradesco Bank in Brazil, which is using IBM’s Watson for serving all internal relationship advisors in a way that they are informed faster and they can go to the end customer lines. To sum up, we save the time of the end relationship manager by leveraging AI internally and then allow them to actually interact directly with the end customer.

And then, in the end, how AI data is used is very important as well. At IBM Watson, our approach is that your data is your data, we don’t actually use the data corpus as an asset that it is monetised. We are focused on improving the AI technology, and not necessarily reusing the data itself.

Artificial intelligence may raise new ethical and legal questions, related to liability or potentially biased decision-making. How does Watson tackle these issues?

At the beginning of 2018, IBM published a Manifesto for usage of the AI, and how data privacy is respected in the context of the projects that we execute. Many people were not aware of its importance until the Facebook Cambridge Analytica scandal happened, however people are now aware. This is really powerful, in terms of raising awareness.

Any additional thoughts regarding AI in financial services? Any future predictions on how you see the market evolving in the next 5-10 years? Will AI be replaced by quantum computing?

If a few years ago people were talking mostly about FinTech, now they have started to pay attention to RegTech and ways to use it for regulatory compliance and fighting financial crime. The next trend is SupTech, or supervisory technology, where you actually help regulators not only to evaluate regulatory compliance, but also to improve surveillance. Recently I was invited to speak to Central banks governors from all central banks across the globe, and SupTech was a key theme that was important for them.

When it comes to the use of AI, people don’t want black box AI; they want to understand the basis on which AI is making decisions, because that could be very dangerous used for example on scenarios of wars, or where ethical questions come into the picture. As a consequence, more transparency is going to be injected into the AI, to avoid usage of black box AI.

And from the quantum computing standing point, you will see a significant emergence of quantum computing being used. Quantum computing is the platform that allows for crunching the types of data that the AI systems need. These processes require immense amounts of computing power that has been unseen in the past. In the last 5 or few years, the fastest super computer in the world was from China, however, the US has finally taken the lead recently based on IBM’s super computer that uses quantum computing tech. Therefore, this will be an emerging trend as well, quantum computing as kind of a platform for AI.

About Vivek Bajaj

Vivek is the Global VP of Solutions for IBM Financial Services. He leads a worldwide team of experienced Sales, Technical Sales and Industry consultants who work with major financial institutions to deliver pre-built industry-specific solutions in the areas of risk & compliance, client insight, payments & counter fraud.

Alongside his leadership responsibilities, Vivek works directly with C-suite executives to share best practices based on his two decades of expertise in financial services. He has personally driven strategic transformation initiatives by leveraging cognitive & advanced analytics at major financial institutions in the recent past. In particular, he has an in-depth understanding of financial services use cases around customer centricity, risk & compliance and counter fraud.