The exploding use of AI in the banking sector

Artificial Intelligence is the buzzword. Everyone is speaking about it, irrespective of whether they are a scholar on the topic or not. But what must be considered is AI's deep-reaching implications and how it has the ability to transform society for the better. In this article, we would look at how AI is currently being used in the banking industry to transform it for the better. It is important to note that while we commoners use the term Artificial Intelligence (AI) everywhere, in reality, it is a mixture of technologies like Machine Learning (ML), Robotic Process Automation (RPA), Predictive Analytics and not to mention, Artificial Intelligence (AI).

Before we dive into the application of AI in the banking industry, let us look into the definition of each term related to Artificial Intelligence, to get a clearer understanding of each term to appreciate their extensive usage in the banking industry.

A)  Artificial Intelligence- In simple terms, Artificial Intelligence is the emulation of human intelligence by computer systems. They can be broadly categorized into weak and strong AI.

Weak Artificial Systems are those who are trained to perform a certain task, and its programming does not allow the AI to learn from its fallacies. The Strong AI or Artificial General Intelligence (AGI), is generally associated with human intelligence and it is responsible for the popular fear in the general population of losing out jobs to increasingly efficient and smooth artificial intelligence.

B)  Machine Learning- Machine Learning is an application of Artificial Intelligence, and as the name suggests, the algorithms enable the computers/machines to learn from its mistakes and mould itself to solve increasingly complex scenarios and problems by taking into account all of its previous data to provide an optimum solution. Machine Learning largely relies on patterns in the data to identify and make better decisions, as and when required.

C)  Robotic Process Automation- To define in layman's terminologies, Robotic Process Automation is the method of combining AI and Machine Learning to process data in high volumes as well as automate repetitive tasks, thereby reducing the time required and human efforts. RPA is largely divided into three categories-

·      Probots- These follow simple rules which are repeatable in nature to process high volume data

·      Knowbots- Bots that are trained to graze the internet to collate much-needed user-specific information.

·      Chatbots- These bots are interactive in nature and are being extensively used in various industries as virtual agents to interact with customers.

Thus from the above pointers, one thing is pretty clear, banks and many other organizations use AI and other complementary technologies to reduce time while increasing efficiency.

Source

Contrary to the global belief that AI with replacing humans, AI will work alongside humans to make the world a better place to live in. The banking industry today has been increasingly using AI in its day to day functions, and it is only expected to rise from here on.

“According to Accenture’s recent Accenture Banking Technology Vision 2018 report, 83% of Indian bankers believe that AI will work alongside humans in the next two years — a higher than the global average of 79%. “93% bankers in India said they increasingly use data to drive critical and automated decision-making.”

The Business Insider AI in banking report states that the front office and the middle office have the largest capability of saving costs through the implementation of Artificial Intelligence in banks. 

“According to Accenture’s recent Accenture Banking Technology Vision 2018 report, 83% of Indian bankers believe that AI will work alongside humans in the next two years — a higher than the global average of 79%. “93% bankers in India said they increasingly use data to drive critical and automated decision-making.”

Source

The above facts in the images have been arrived at after analyzing the following financial institutions/companies- Citi, JPMorgan Chase, U.S. Bank, Personetics, HSBC, Quantexa and Capital One.

AI is benefitting all sectors of business, and banking is no exception to this. Some of the recent uses of AI in the banking industry are enumerated as follows-

i)               Personal Assistants and Chatbots- In this era of digitization, companies need to not only leverage mass digital marketing but also customize their products and services for each customer. AI helps in individual customization through the use of chatbots and personal assistants. Customers can now not only choose a product of their liking but also leverage these platforms for redressing grievances, contacting customer support and writing reviews to name a few.

Mobile banking, another successful effect of digitization, has been integrated with Artificial Intelligence to yield wonders for the banks.

"As per the survey by the National Business Research Institute, over 32 percent financial institutions use AI by the means of voice recognition and predictive analysis."

Source

Today's knowledgeable customer uses mobile banking a lot as it is easier, faster and efficient. Thus, banks can get detailed information about their customers through predictive analytics based on their mobile app usage, and they can then focus their attention on the products and services that are most liked.

ii)               Backend automation using Artificial Intelligence- Customer data and critical image documents (like Aadhar, Voter Card, Pan Card) can be processed extremely fast using Artificial Intelligence and Machine Learning to yield valuable information about the client(s). Robotic Process Automation can make this work easier by extracting only the valuable lines of information from these documents and analyzing them to create a complete profile of every customer the bank deals with, ever.

iii)             Security- AI can be successfully implemented in the middle office to detect fraud and money laundering. An Artificially Intelligent system can successfully detect these anomalies and report them to the higher management in the blink of an eye. Better fraud and money laundering activities detection will lead to better management of financial resources within the bank thereby leading to not only greater profits, but it would also mean that more people would have access to the funds of the banks (as loans) thus helping the economy as a whole.

“Most banks are looking to deploy machine or deep learning and predictive analytics to examine all transactions in real-time. Machine learning can play an extremely critical role in the bank’s middle office.”

Source

iv)               AI in ATMs- One of the most necessary places where Machine Learning and Artificial Intelligence is the need of the hour. Real-time facial recognition software supported by robust artificial intelligence and machine learning would lead to the lessening of the incidences relating to ATM vandalism. India has already witnessed a lot of loss because of ATM thefts and vandalism and at times surprisingly pests (rats) chewing away through wads of cash inside the ATM storage locker!

“Bank heists have not gone out of fashion just yet, both as a cinematic trope, and also as a shortcut to riches. A total of 972 such incidents were reported in 2017-18, roughly three every day, according to data collated by the Reserve Bank of India (RBI). The banking sector lost a total of Rs 168.74 crore to organized crime directed at ATMs in the past three years. This includes figures for the first quarter of FY19. Between April and June 2018, 261 incidents were reported, entailing a loss of Rs 18.85 crore to banks.”

Source

v)               Trading and Securities- Robotic Process Automation (RPA) is extremely useful in trading and securities; an RPA could easily distinguish between the prices for the purchase of securities between an investment manager and a broker and then create an email message directed towards the broker on predetermined rules. RPA can also streamline several processes where one would need to skim through rows and columns of data to arrive at the correct financial decision.

vi)               Portfolio Management- Machine Learning, equipped with years of data on the stock market of various stock exchanges throughout the world are now able to construct near-perfect portfolios that minimize the risk all the while aiming to maximize the wealth over a period of time.

Challenges of implementation of AI

Any technology would have its fair share of challenges in implementation, and AI is no exception in this regard. AI without data would be akin to a human without a heart and a brain. Also, the availability of the right data is another issue. Hence, there needs to be a structured process for collecting customer data and later, cleaning, processing, and finally analyzing the same to reveal the important insights that banks need so much. Then comes the additional burden of the language barrier in India, with tens of different languages and hundreds of dialects spoken, the process of collecting regional data would be a herculean task for sure. Banks from all over the country needs to come together to form a framework and a structure so that these challenges become easier to face. AI has the potential to become a trillion-dollar industry by the year 2035 and its explosive growth in the present times only strengthens the statistics.

In addition, government banks must also be willing to partake directly informing the framework mentioned, so that the poorest of the poor in the economy benefit from these technologies. After all, no development can really be termed as such if it is not inclusive.

References-

https://www.businessinsider.com/the-ai-in-banking-report-2019-6?IR=T

https://www.livemint.com/AI/v0Nd6Xkv0nINDG4wQ2JOvK/Artificial-Intelligence-in-Indian-banking-Challenges-and-op.html