2020 AI Trends in Banking

2020 Lockdown Loading
AI in Banking Trends

Trend #1: AI in Banking

Trends in Banking in 2020 and AI since the COVID-19 lockdown: 

The concepts of money and transactions have seen it all! It started with the barter system where people traded goods or services in exchange for a different type of goods or services. It was eventually replaced by currency, which then became the standard norm across the globe wherein people paid or received cash in exchange for goods or services. Fast-forward to the 21st century when cards and smartphones have replaced the need to keep cash in your wallets. With the swipe of a card or the scan of a barcode with a smartphone, people are now able to perform transactions within a blink of an eye and without the hassle of counting cash and coins. As is evident, the notions of banking and transactions have come a long way. In this article, we look at how the advent of technology and artificial intelligence has been further exploited to give an entire new dimension to the banking spectrum.  

Banking Channels and Artificial Intelligence

To discuss the utilization of Artificial Intelligence in banking, we first look at the three main channels of the banking industry:

ai robot
Pepper Robot by Jason Margolis/ The World 
  • Front office: Earlier, people had to walk in physically to the bank to open their accounts, deposit money, take money out etc. Even for small inquiries, they had to call the bank at the least. This has now gotten replaced with the introduction of applications and websites wherein customers can easily open their bank accounts on their smartphones/laptops.Further, to enable a two-way customer experience and enhance a conversational atmosphere, banks have now seen the introduction of chatbots on their respective websites and apps. A latest report by Gartner suggests that by 2022, 70% of the white-collar workers will interact with chatbots on a day-to-day basis. In addition, Voice Banking is considered by many as the next revolutionary step in customer interaction. Some European and Japanese banks already have a humanoid robot named ‘Pepper’ which are accelerating the digital transformation among financial institutions
  • Middle office: Conventional case management systems for fraud detection wherein the lenient rules-engines generated a huge number of false positives, have been replaced with robust machine learning based Anti-Money Laundering and Anti-Fraud models. These are not only more accurate but also reduce the generation of false positives, thus presenting the banks with a great cost savings opportunity.
  • Back office: The process of collecting and reviewing customer documents manually has now been conveniently replaced with AI which has incorporated scorecard development models to generate the credit score of the customer within minutes; and instantaneously make the decision as to whether he/she should be granted the loan. 

We now look at the AI technologies in the AI Wheel above that will disrupt and change the face of the banking and many service sectors in 2020. 

Trend #2: Convergence     

Data Convergence Technologies

Blockchain and Quantum Computing

Flashback to the time when every person on the planet was trading bitcoins! Bitcoin was built on the concept of tokenizing the traditional forms of monetary instruments and setting them up on public blockchains. Blockchain has the power to remove intermediaries in the payments, KYC onboarding and loan setups, at the same time ensuring that the transactions are tracked in a secure and verifiable way. Ameliorated transaction and customer satisfaction imply that the prospects of blockchain are huge when it comes to the exchange of money and data. 

A great example of AI in blockchain can be in the credit lending sector where most firms are obliged to use simple models like logistic regression which are easily explainable. This is because in complex models such as deep learning models, it becomes much harder to determine why a certain decision has been made. With blockchain, we will be able to record how separate actions result in the final decision which allows us to go back and tweak the model. A fintech AiX negotiated its first Bitcoin-USD trade in February using a proprietary AI system that communicates with human traders at banks via chat services; and became one of the top AI-Powered cryptocurrency Trading Bot in 2020.

That being said, quantum computing has emerged as a fierce competitor and it is widely understood that it could replace blockchain within the next ten years. It uses the concepts of superposition and entanglement to create Qubits, and this quantum encryption allows the banks to send data over a quantum network. The main benefit of quantum computing lies in the power of encrypted messages, because of which the data cannot be hacked. Any attempts at hacking or tampering them leads to their automatic and immediate destruction, thus making quantum computing such a powerful force. Barclays and JP Morgan are among the few banks that have already started experimenting with this concept.

Trend #3: Automation in 2020:

Robotic Process Automation (RPA)

Similar to any other industry involving technology, banking also involves a host of repetitive and manual tasks. These include opening applications, sending and retrieving emails, data reconciliation, copying information from one database to another etc. Due to the manual nature of these tasks, they are labor-intensive and prone to errors. This is where RPA comes in handy. RPA is basically the development of robot assistants which perform these repetitive tasks according to the way they are programmed. Using RPA processes, developers can now concentrate on complex and innovative tasks, therefore driving the economic potential of the financial firms. 

Trend #4: Cloud Frontier in 2020

Cloud Computing

There is only so much data internal databases can store and handle! This is where cloud computing comes into picture. This technology not only stores humongous quantities of data on the internet, but also delivers services such as servers, databases and analytics. With new machine learning models getting into the picture now more than ever, cloud computing also enhances the scaling up of these models on the large number of customer data points banks now have. In that respect, many firms have already incorporated cloud-based software-as-a-service (SaaS) applications in a lot of their business units. 

Trend #5: 2020 Security

Cyber Security 

Banks and financial service companies handle a lot of personal and sensitive information. This becomes a lucrative target for cyber frauds who try to hack in the systems and use this information for illegal purposes. The security of this data is at a tremendous risk given rapid growth of technologies and use of mobile applications by customers, most of who indulge in cashless cross-border transactions. In case of a mishap, not only do the customers lose trust in the bank, but the banks have to pay millions of dollars worth of penalties to the regulators. In this respect, cyber security is the need of the hour. Cyber security entails the use of security audits, firewalls, anti-virus, multi-factor authentication and biometrics to fight this cyber war. 

Trend #6: 2020: AI for IT Operations (AI Ops)

AI OPs or ML OPs

AI Ops, through machine learning, is constantly and repeatedly learning patterns and improving IT and DevOps processes. Model monitoring becomes especially important when these models and solutions involve monetary risks. Let’s take an example of the stock market and the products revolving around this interesting topic. 

Years ago, the stock market was driven by telephonic calls and messages. As soon as we got access to fast and cheap internet, the medium of trading moved online. In the last decade or so, algorithmic trading has revolutionized the trading scenario. However, predicting the movement of stock prices, foreign exchange, bonds, options and commodities such as gold, oil etc. has always been and is still a tricky business. With the advent of machine learning and deep learning algorithms such as Recurrent Neural Networks, this has been even more competitive and cut-throat. Scientists are now beginning to move beyond the conventional numeric datasets and targeting twitter sentiments to predict these movements. This is where the need to implement tested and profitable models comes in. Read more about how a financial institution gained a competitive edge in the bonds market using Datatron’s products. Datatron not only has robust inbuilt algorithms but their Machine Learning Model Operation (MLOps) platforms ensure proper model management at the production and deployment level. 

In conclusion, we await an exciting period of time in 2020 and beyond, wherein technology and artificial intelligence is expected to be used insanely to come up with new innovations and enhance the existing processes.



Let's Discuss