November 26, 2019 – A report from the World Economic Forum in 2018 said that over 75% of banking leaders considered artificial intelligence (AI) to be a top priority (https://www.fintechfutures.com/2019/11/what-does-ai-mean-for-the-banking-industry/) if they were to differentiate themselves from the market. Whilst a statement from the CEO of Deutsche Bank two years ago terrified many workers when he stated robots could replace half of the staff jobs (https://www.cnbc.com/2017/11/08/deutsche-bank-ceo-suggests-robots-could-replace-half-its-employees.html), thankfully AI has not been so apocalyptic and has many practical applications.
Many large banks and some smaller ones are now deploying AI solutions. The systems used are varied but there are certainly cases that are more mature than others. In this article, we will look at some of the key AI deployments in the banking sector.
Why do banks want to use AI?
Customers have become tech-savvy and expect their banks to deliver the seamless experience they get at Amazon, Netflix or other retail and entertainment sites. To enable this, the financial sector has invested quite heavily in digital technology like mobile applications and e-banking. Customers can now deal with their banking services whenever they need to, without having to visit a branch or make a phone call. However, some might argue this has come at a cost.
As so much personal and financial data is being transferred between mobile devices and cloud servers, banking has become more open to malicious attacks. That, in turn, has led to tighter regulations as we’ve seen the rise of new Fintech players such as Monzo coming to the fore with the capacity to be more nimble than established institutions.
In order to stay competitive, banks have turned to (AI) for innovative solutions.
Chatbots and Customer Service
Conversational chatbots have started to become a significant part of the banking industry. The customer experience with banking has historically been left alone, causing customers to have long waits in branch or spend time on the phone attempting to talk to an advisor. A chatbot is the perfect AI solution to solve the problem.
A chatbot is a digital support bot that can talk to customers in a conversational interface without them needing to speak to the bank directly. There are several examples used in banking such as that deployed by the Bank of America.
The Bank of America chatbot has introduced Erica (chatbots are often given humanized names, https://promo.bankofamerica.com/erica/), which sends notifications to customers, provides information about their balance and suggestions on how to save more money amongst other things. Since launch, Erica has been developed into a virtual assistant that helps clients make better-informed decisions. All of this is done via the mobile banking app.
JPMorgan, Wells Fargo, Capital One, Ally Bank, HSBC (Hong Kong) and SEB (Sweden) are some of the other banks that have successfully deployed chatbot technology in recent years. Whilst they have different applications, the objective is always to improve efficiency and customer experience. Lloyds Banking Group is using its bot to help employees better with service rather than being directly customer-facing.
Fraud Detection and Money Laundering
This is arguably the most important application of AI in banking. If those in the sector are going to invest in technology and mobile applications, it opens them up to those wanting to cause malicious damage. Every new technology added to a bank gives another route for hackers to access the institution.
Danske Bank has modernized its fraud detection capability using a technology provided by Teradata (https://www.teradata.co.uk/Press-Releases/2017/Danske-Bank-and-Teradata-Implement-AI). The AI system is using machine learning applications that can quickly pick up on anomalies across transactions. Over time, it learns from any mistakes it makes and becomes more intelligent. As a result of the implementation, Danske has increased the detection of real fraud by 50%. However, just as importantly they reduced false positives, those that were previously believed to be a fraud but were not, by 60%. In reducing false positives, Danske Bank can focus on genuine cases.
There are many vendors offering similar types of technology (https://emerj.com/ai-sector-overviews/artificial-intelligence-fraud-banking/. Feedzai, for example, offers anomaly detection-based fraud software that develops detailed risk profiles on customers, scoring them against very granular data. They claim to have helped a top retail bank in the US more accurately determine fraud.
Predictive banking is another way for banks to enhance the customer experience. More often than not, it is tied to the use of mobile apps and bots.
According to Deltec Bank, “The concept of predictive banking is to allow customers to understand how much money they will have at the end of the month and not just focus on what they are spending now. Historical data will be used to predict future trends and events.” The sophisticated AI algorithms can process huge amounts of information and determine what a customer’s position is likely to be if they complete a specific action.
As well as a customer benefit, it might help the banks by reducing costs if the customer does get into debt.
As an example, Wells Fargo has a predictive banking app (https://emerj.com/ai-sector-overviews/ai-in-banking-analysis/) that alerting customers if a bill payment is higher than average, tells them to set up a travel plan on their account if they buy a plane ticket or reminds them to send money to savings. The application has 50 different prompts that can be used depending on the scenario.
Robotic Process Automation (RPA)
Banks are relying on bots to reduce their costs through operational efficiency. Whilst automation is not necessarily AI (this has been a subject for debate amongst experts), the software often needs to integrate with some kind of intelligence.
For example, the Bank of NY Mellon Corp has worked with technology company Blue Prism to roll out over 200 bots (https://emerj.com/ai-sector-overviews/ai-in-banking-analysis/) that can handle repetitive tasks. This includes things like transferring funds or responding to external auditors. One such task could be reviewing data to correct formatting and mistakes which would take a human far longer to complete.
In adding this machine learning to the banking tasks, BNY Mellon has reported an 88% improvement in processing times and a 100% accuracy in account closure validations.
Future of AI in Banking
The applications above are great examples of how banking is moving into the 21st century. Fraud detection, chatbots, RPA and predictive banking are all becoming standards across most financial institutions to allow them to remain competitive.
In this article, we haven’t talked about potentially revolutionary technologies like Blockchain and Open Banking. These will undoubtedly transform the sector as we know it but are still gathering momentum but probably required articles in their own right to fully explain what they could do.
The case studies provided show how the banks are using AI to forge an advantage and we will continue to see this moving into 2020.
Disclaimer: The author of this text, Robin Trehan, has an Undergraduate degree in Economics, Masters in international business and finance and MBA in electronic business. Trehan is Senior VP at Deltec International www.deltecbank.com. The views, thoughts, and opinions expressed in this text are solely the views of the author, and not necessarily reflecting the views of Deltec International Group, its subsidiaries and/or employees.
About Deltec Bank
Headquartered in The Bahamas, Deltec is an independent financial services group that delivers bespoke solutions to meet clients’ unique needs. The Deltec group of companies includes Deltec Bank & Trust Limited, Deltec Fund Services Limited, and Deltec Investment Advisers Limited, Deltec Securities Ltd. and Long Cay Captive Management.