The Evolution of Banking: AI
What is AI?
The concept of Artificial Intelligence (AI) was first coined in 1955, by Stanford Professor John McCarthy, as ‘the science and engineering of making intelligent machines’ . In other words, it is a branch of Computer Science, in which a machine mimics the cognitive functions that are associated with the human mind – such as learning and problem solving. Some famous examples include high-level strategic gameplay (e.g. when Google’s DeepMind beat the world’s best Go player ) and autonomous cars. For machines to act and react like humans, they need access to sufficient information from the real world. To be able to initiate common sense, reasoning and problem-solving power, AI needs knowledge engineering – which provides data about objects, categories, properties etc. . The graph below shows the different branches of AI .
Exploring Machine Learning
Following on from this is another buzzword: Machine Learning (ML). This is an application of AI, based around the idea that we should let machines learn for themselves. With the rise of the internet, there is a lot of digital information being created – which means there is more data available for machines to analyse and ‘learn’ from . For this to be successful, humans need to hand-engineer features for the machine to look for. Delving deeper into ML, deep learning (DL) is all about reducing the human engineering aspect. One of the most common approaches to this is artificial neural networks, which is a mathematical system based on how neurons work in the human brain. DL focuses on making the machine classify information in the same method as the human brain. By working on a system of probability, the machine makes decisions, statements and/or predictions with a specific level of certainty. This is entwined with a feedback loop system, which regularly tells the machine if a decision made was right or wrong .
According to a report from the consultancy Accenture, Artificial intelligence will be the main way that banks interact with their customers within the next three years . There is a big misunderstanding that AI will automate the banking process, and result in a less personalised experience for the customers. In fact, the opposite is true. AI will allow banks to better understand their customers, and analyse far deeper than a human would.
J.P. Morgan Chase recently introduced a Contract Intelligence (COiN) platform, designed to analyse legal documents and extract the important points/clauses. One example of how they used this was for the review of 12,000 credit agreements. This would normally take around 360,000 human working hours, but the AI platform finished the job in a matter of seconds .
In 2016, Bank of America launched an intelligent virtual assistant called Erica. This chatbot aims to use predictive analytics and cognitive messaging, to provide customers with personalised financial guidance – to reach their financial goals. The assistant would also be able to perform day-to-day transactions and be available out of banking hours, 24/7 .
CitiBank’s investment and acquisitions division, Citi Ventures, invested into Feedzai – a data science firm that can identify and eradicate fraud in real time. By conducting large-scale analysis, and detecting any questionable customer actions, it can monitor potential threats at greater accuracy and speed than humans can .
BNY Mellon is using AI at a high level, to reduce costs and eliminate any tasks of repetitive nature. Over the last 2 years, they rolled out over 220 bots (created by Blue Prism) to process automated tasks. This includes account closure, trade entry, transferring funds, responding to data requests from external auditors, correcting data mistakes/formatting and much more .
The Future of Banking
A 2017 report on banking trends highlighted enhanced customer personalisation as the number one trend . With AI, algorithms can offer portfolio management advice, that is regularly updated based on the customer’s data. It will also allow banks to offer targeted offers and better market specific, relevant products. For investment houses (hedge funds, prop trading etc.), algorithmic trading can be enhanced by using AI to analyse the market sentiment. In fact, some reports claim that over 70% of trading activity today is undertaken by AI systems .
There are also productivity gains to be made, since repetitive and back office processes can be automated. Especially with the financial services industry facing stricter and stricter compliance requirements, AI can be used to evolve and work around new regulations. For instance, AI can be very useful in AML (Anti-Money Laundering) pattern detection – by moving away from rule-based software to a less rigid search list.
The machines won’t take over – yet. AI will gradually replace humans in many fields, but the risks of bias, privacy, trust etc. will create many obstacles along the way. With many banks facing falling profit margins, increasing customer expectations, and increasing competition from FinTech start-up’s – they need to reduce costs and improve their offering. It will be interesting to see how the face of banking will change over the coming years, especially with many tech firms (Apple, Facebook, Google etc.) heavily investing into AI and financial services.
Users who have LIKED this post: