Page 48 - Banking Finance January 2022

ARTICLE

on now is data acquisition. Lack of proper customer records
is the biggest obstacle to AI.

We need to make sure that the data used by banks are KYC
compliant clean data as it is used in AI models. Huge data
infrastructure is needed to impact AI. Verification of the
correctness and accuracy of data is also required before
using such technology in the public domain.

Analysis and authentication of data:
The amount of data with the banks is very large, while
Oracle and Accenture store all the data in the bank. What
we need is a proper analysis of the data, and it requires a
high level of leadership expertise to bring cross-functional
Explaining it to regulators can be a challenge for teams together - one with knowledge of financial business,
bankers. The Justice Srikrishna Committee said that the and the other with different disciplines for the effective use
biggest challenge in using big data and artificial of such data sets with the required machine learning skills
intelligence is to work outside the framework of to create a framework and infrastructure. AI remains a
traditional privacy principles. It now operates in the niche-based domain with a shortage of talent and expertise.
reverse way and risks banks unknowingly. This leads to
hidden bias in decision making as AI has access to all Leakage and data misuse
users 'data. U.S. And the U.K. Many experts in cybersecurity believe that
cyber, political and physical threats arise with the
3. Reduced Customer Loyalty capabilities and reach of AI. The recent Facebook scandal
Y There is also a fear that a lack of customer contact and highlights the risks that corrupt data practices can bring to
a lack of the essence of "human touch" will diminish a company. We also need to ensure complete transparency
customer loyalty. Because banks, especially in India, help when entering new AI projects so that banks do not face
so many people in fulfilling their long-term dreams, they reputation risks.
have emotional value - be it a beautiful home or a good
education for students. All of this could be lost due to Banks should start building AI SYSTEMS with small complex
AI and automation. Socio-economically disadvantaged data and add the latter, thus creating a universal record of
groups are the biggest losers and are most affected by each client. Adequate investment should be made in storing
such a low level of education and the digital divide. data securely and preventing leakage. This helps the bank
to identify potential risks during the project implementation
Future Action phase and to effectively identify and then implement the
company's goals and priorities. Artificial intelligence will soon
Tech columnist Nick Bilton writes in the New York Times,
"[Insurgents of artificial intelligence] grow quickly and become the sole determinant of banks' competitive position
become scary and catastrophic. Imagine that the medical and is the key to maximizing their competitive advantage.
robot, which was originally programmed to eliminate
cancer, was able to conclude that the best way to eradicate Summary
cancer is to eradicate infected humans. "The message here Artificial Intelligence (AI) is a science and a set of
is that banks need to raise awareness of the impacts of computational technologies that are inspired by-but typically
banks. Develop digitization and broad foresight in the operate quite differently from-the ways people use their
possibilities of AI - so that we as humans have control over nervous systems and bodies to sense, learn, reason, and
AI and not the reverse. One area that banks should focus take action. While the rate of progress in AI has been patchy

48 | 2022 | JANUARY | BANKING FINANCE