Optimisation of the reception and engagement of clients
Understanding the offerings of financial products and services is already greatly facilitated by banking assistants and “online” simulators, such as chatbots or voice bots. For example, Erica, the chatbot for Bank of America, can answer the questions most often asked by customers-it then becomes much faster to request payments, view account balances, search for past transactions. JP Morgan’s virtual assistant is the first AI program in the world of corporate payments and is particularly compelling.”Based on your behaviour each time, it will start to learn what you ask for,” Jason Tiede, Innovation Head for Treasury Services, said in an interview with CNBC. Furthermore, within HSBC, Amy can provide instant support to customers’ inquiries on a 24/7 basis customer servicing platform.
In retail banking, asset management, insurance and investment banking, product recommendation is an essential activity, requiring the detailed exploitation of the information available on the client. A set of data made up of a history of payments, messages sent and received, emails and activity on social networks allow banks to better target customers regarding consumption habits, the chances of a full refund of a loan etc. AI also makes it possible to directly connect lenders to borrowers on new credit platforms and other third-party providers. As a result, customer credit application processing times are significantly reduced to their satisfaction. Besides, this particular knowledge of customers allows “automated advice”. Robo-advisors use the client’s profile to allocate a portfolio compatible with their risk profile. They can also advise the client, depending on market developments, to make specific investment decisions.
The almost instantaneous processing of request forms or customer feedback allows the development of new services and increases the satisfaction rate.
Fight effectively against fraud and money laundering
According to a study conducted in 2017 by the Observatory for the Security of Payment Means, no less than 5.1 million fraudulent cases were noted, equating to more than 744 million euros.
Signatures, confidential code, SMS or verification email, are some examples of tools that have been implemented to protect customers from fraudulent use of their bank card. But that is still insufficient today. Thus, banks’ goals are to integrate machine learning into their system to analyse their customers’ banking behaviour to spot a suspicious transaction immediately. Machine Learning is one of the fields of AI study, allowing algorithms to learn independently, based on data provided to them, to make decisions or solve problems tasks without being programmed for this purpose. After having “created customer profiles”, updated regularly, AI can predict or detect a fraudulent transaction based on the customer’s usual, expected behaviour to detect abnormal behaviour. It can also be found on other techniques such as behavioural biometrics, which considers the standard way the customer’s phones and tablets are used. By analysing his typing force on the screen, his scanning angles, his typing habits on a keyboard, it becomes much easier to detect something abnormal.
For instance, Mastercard has developed a solution integrating AI called Smart Agent to detect any fraudulent transaction. However, it is no longer enough to detect fraud – a quick and effective response is also necessary. AI enables automatic and instant response to various frauds. Danse Bank, a Danish bank, has developed an AI-based solution capable of analysing transactions in “less than 300 milliseconds”. By becoming the first bank to implement an AI-based platform combining Deep Learning and Machine Learning, Danse Bank can automatically remedy fraud by limiting false alarms.
The fight against fraud is a crucial issue for banking establishments. The deployment by banks of AI-integrated solutions to enhance transaction security, in turn, builds customer confidence.
However, detection or response solutions can pose some problems. First, these can generate false positives. In addition, it is essential to note that behavioural analysis poses an ethical issue in developing a behavioural model of a customer by their bank without their consent.
Strategic and decision-making tools
AI and machine learning not only allow optimisation of the customer experience but also revolutionise the banking sector’s business processes.
In the market activities of investment banks and hedge funds, a significant part of the trading occurs in algorithmic form. High-frequency micro-arbitrage between quotation platforms allows fast and efficient transactions.
Currently, these algorithms, although ultra-fast, do not learn from their environment throughout exchanges. This creates great instability – for example basic stop-loss rules. Once again, machine learning would be ideal for algorithmic trading.
In addition to automating repetitive tasks, artificial intelligence can replace financial experts in certain situations. In the context of financing requests, AI can analyse banking history, professional status, income, or even the clients’ wealth precisely and rapidly. These diagnostics substantially help financial advisers who can then make decisions quickly while minimising risk-taking.
Our society is in the process of transformation, and artificial intelligence and its technologies allow us to keep a clear, transparent and delivered customer promise. By enabling a large amount of available customer and market information to be analysed in real-time, the best solutions can be predicted and implemented. Today, the banking profession once again takes on the status of a banking expert. Assisted by AI and released from low value-added activities, bankers can effectively listen to their customers and offer them the products that best match their situation and demands. According to a McKinsey report ,ensuring the adoption of AI technologies across the enterprise is no longer a choice, but a strategic imperative. Whether it is to improve the quality of services, customer relations or optimise portfolios and financial services, AI is a significant asset for banks. Will banks be prepared to meet the challenge?
Written by Hélène Pignon