Page 47 - Banking Finance January 2022

ARTICLE

the customer's creditworthiness. It helps in giving pre- game-changer by detecting insider trading that leads
approved loans to a large range of customers without to market abuse.
the need for paperwork and allows the self-employed
and students (because they lack the financial fold) to 8. Insurance Underwriting and Claims:
get credit. In the case of SME and corporate lending, Y In this era of bankruptcy, consumers are more likely to
AI simplifies the complex borrowing process, analyzing arrive at banks than visit insurance companies. The
market trends and identifying potential risks in lending, insurance industry can take advantage of AI in detecting
future behavior and even the slightest likelihood of underwriting, claim-handling policies and fraud. This
fraud. helps identify risky behaviour and charges higher
premiums to groups of customers. There is an enormous
7. Risk management and fraud detection amount of data in insurance companies that can help

Y The Punjab National Bank scandal has put the banking you create mathematical models and accurately predict
sector at enormous risk, shaking up regulators, financial risky behaviours. Banks can also provide such data for
and stock markets and the banking industry. AI and use in customer risk identification. This reduces turn-
appropriate due diligence can monitor such potential around-time (TAT) for both loans and insurance. For
threats and help banks to install fool-proof surveillance example, in order to analyse the damage to the vehicle,
and fraud detection systems. Surveillance in banks was deep learning techniques can analyse the image of the
done through audits and sampling. Some data sets and vehicle and calculate the cost of repair using the models
files that cause major losses are not covered in these attending.
models. The algorithmic rules-based approach helps
monitor each file, and machine learning techniques can Threats by AI
keep a database of such files vulnerable to the bank.
Alibaba founder Jack Ma warned viewers at the World
Y Banks can use AI to detect fraud in transactions or to Economic Forum 2018 in Davos that AI and big data are a
detect any suspicious activity in a customer's account threat to humans and will stop people from empowering
based on behavioral analysis while providing safe and them. The massive expansion of AI in banks comes with its
fast transactions. With increasing cybercrime in recent share of risks and opportunities. Banks increase their
years, AI can be used to manage cyber-security and, investments in AI every year, often with obsolete risk. We
most importantly, protect personal data. Citibank has need to understand the risks that the AI can also face.
already invested over $ 11 million in new money
laundering, using machine learning and big data. 1. Loss of jobs
Y AI-based systems help in compliance by ensuring the Y Automation of Tasks Banks faces the risk of backlash
functionality of internal control systems. AI is also a from their employees, leading to job loss and job
restructuring. AI, in a way that maximizes
organizational productivity, redefines the way
employees do their jobs. This can lead to dissatisfaction
among employees, resulting in resignations or layoffs.
AI can also replace a teller, customer service executive,
loan processing officer, compliance officer, and finance
manager.

2. Process Opacity:
Y Although deep learning models and neural networks in
AI have proven to be more complete than human
decision-making over time, they are often not
transparent in terms of how such conclusions are made.

BANKING FINANCE | JANUARY | 2022 | 47