The Legal Aspect And Evolution Of AI In Global And Indian Banking
Author: LawSphereX
Category: Technology, Cyber & Privacy
Published:
Updated:
THE LEGAL ASPECT AND EVOLUTION OF AI IN GLOBAL
AND INDIAN BANKING
Shivam Swaroop Mathur, BBA LLB, Symbiosis Law School, Nagpur
ABSTRACT
This paper examines the transformative influence of artificial intelligence (AI) in the banking sector globally and in India, focusing on the legal, ethical, and practical dimensions of AI integration. As banks harness AI for enhancing efficiency, personalising customer service, managing risk, detecting fraud, and offering financial advice, deploying AI technologies raises significant questions regarding privacy, accountability, regulatory compliance, and the potential displacement of human labour. This study delves into the benefits and challenges of AI in banking, scrutinises the legal frameworks regulating AI use in various jurisdictions, with a particular focus on India, and contrasts these with international standards.
Through this comparative analysis, the paper aims to illuminate the evolving landscape of AI in banking, emphasising the need for robust legal strategies to harness AI's potential while safeguarding against its risks.
INTRODUCTION
Artificial intelligence has emerged as a transformative force in the banking sector worldwide. As financial institutions grapple with increasing data volumes and the need for efficiency, AI technologies offer promising solutions. In this context, we explore AI adoption's legal implications and evolution in global and Indian banking systems. The convergence of AI algorithms, big data, and computing power has led to applications that range from personalised customer service to risk assessment and fraud detection. However, these advancements come with critical questions about privacy, accountability, and regulatory compliance. Understanding the legal landscape becomes paramount as banks integrate AI into their operations. This project will explore the multifaceted relationship between AI, banking, and the law. We examine AI deployment's benefits, challenges, and ethical considerations.
Additionally, we analyse the legal frameworks governing AI in India and compare them to international approaches. By doing so, we aim to shed light on the evolving landscape of AI in banking and its legal implications.
APPLICATIONS OF AI IN BANKING
AI will enable financial companies to rethink their business models, provide cutting-edge goods and services, and significantly impact customer experience interventions. In the second machine age era, banks must contend with emerging fintech companies utilising cutting-edge technologies to supplement or replace human labour with complex algorithms. Banking companies must embrace and incorporate AI into their business plan to keep a competitive edge. The areas where the AI-enabled technology investigated the modifications include the following ones. 1 • Risk Assessment AI replaces human analysts in corporate operations because selecting people manually is expensive.
Artificial Intelligence can automate processes in domains where needed, provide intelligent analytics, and foster clear thinking because it learns over time and can guarantee the highest accuracy in calculations and data analysis. As a result of their demonstrated effectiveness as a customer service tool and the time and money they save 1 Dhanabalan, T. , and A. Sathish. "Transforming Indian industries through artificial intelligence and robotics in industry 4. 0. " International Journal of Mechanical Engineering and Technology 9. 10 (2018): 835-845.
businesses, AI-powered chatbots appear to be an unassailable resource. This cuttingedge technology would transform both non-commercial and commercial operations. 2 • Fraud Detection and Management Reducing the risks linked with the business is the ultimate goal of any enterprise. This assertion would be appropriate for financial and banking organisations as well. Banks and other financial organisations must quickly receive and disburse interest on deposits and dividends since they handle deposits and payments. For this reason, financial institutions and banks view fraudulent activity as a severe problem and use AI-enabled tools to prevent financial pollution. 3 Artificial intelligence is the best technology for fraud and security detection; It can evaluate a person's past spending patterns and behaviour toward different transactions.
From there, it can spot unusual behaviour, such as using a card from a different nation just hours after it has been used elsewhere or if there has been an extraordinary attempt to withdraw money from an account under investigation. Another stunning feature of AI fraud detection is its unwavering confidence in experience-based learning. When a person corrects a typical transaction that raises a red flag, artificial intelligence allows the system to learn from the experience and make informed conclusions about what constitutes fraud and what doesn't. 4 • Financial Advisory Services The PWC Report from 2017 discussed the potential for robot advisors shortly. Technologies will do tasks humans cannot complete for a single down payment due to the pressure placed on banking and financial institutions to reduce the commission amount on individual investments.
In addition, a novel approach known as bionic advice is employed in this procedure. Combining mechanical and human ideas results in highly efficient alternatives beyond what the separate components could provide. 5
2 Techlabs, Maruti. "Why can chatbots replace Mobile Apps immediately." Retrieved March 9 (2019).
3 Ganapathy, S., and Mr S. Rajamohan. "Technology adoption in public sector banks—An employee-centric study." DogoRangsang Res. J 10 (2020): 345-355.
4 Techlabs, Maruti. "Why can chatbots replace Mobile Apps immediately." Retrieved March 9 (2019).
5 Anantharaman, Ashish. "How artificial intelligence is enabling financial inclusion in India." (2019).
Trading Investment firms rely increasingly on technologies and data analysts to predict future market positions. The ability of investment organisations to accurately predict the market's future is critical to their success. Systems significantly impact this forecast since they have quickly processed vast amounts of data. In this manner, systems that have been taught to analyse different kinds of historical data can forecast how the data may recur. While this happens, abnormalities related to the 2008 financial crisis are present in the data. The companies' systems are trained to analyse the data, identify the causes of these anomalies, and develop strategies to prevent them from occurring in the next prediction. 6 Due to its greater reliance on personal risk tolerance, AI may offer various answers to meet everyone's needs.
Thus, an individual with a higher risk tolerance can leverage artificial intelligence to solve different problems related to buying, holding, or selling equities. Conversely, those with lower risk tolerance can utilise AI-powered tools to receive alerts about potential market dips or rises. As a result, businesses may decide to increase their investment, hold onto their current position, or exit the market. • Automated Customer Support These days, AI-enabled technology like text chats, voice systems, and chatbots have supplanted the conventional customer care sector. With the help of this cutting-edge technology, clients may now receive executive-level customer support and expert advice at a reduced cost. AI-enabled solutions have thus taken the role of the conventional customer service processes previously in place, saving businesses time and energy.
7 • Claims Management AI-enabled systems collect data and intelligently evaluate it to guarantee high-quality outcomes. AI-enabled solutions pick up on users' or customers' unique behavioural 6 Dhanabalan, T. , and A. Sathish. "Transforming Indian industries through artificial intelligence and robotics in industry 4. 0. " International Journal of Mechanical Engineering and Technology 9. 10 (2018): 835-845. 7 Malali, Anil B. , and S. Gopalakrishnan. "Application of artificial intelligence and its powered technologies in the Indian banking and financial industry: An overview. " IOSR Journal Of Humanities And Social Science 25. 4 (2020): 55-60.
patterns to identify anomalous activity and notify users of anomalous transactions and situations. These technologies are crucial to the administration of claims; in particular, machine learning techniques enable the omission of numerous steps in the claim settlement procedure. 8 • Accurate Decision-Making In the era of data-driven management, businesses like banking, finance, and insurance will benefit from more affordable management decisions. These organisations will direct their inquiries toward the systems rather than the experts. By evaluating the vast amount of data, systems and machines would produce the desired outcomes, aiding in the decision-making of corporate managers and their subordinates.9
CHALLENGES AND BENEFITS OF AI IN BANKING
The potential for innovation and efficiency in Indian banking is enormous when AI and ML are integrated. Nonetheless, careful consideration must be given to the ethical and legal ramifications. The existing regulatory environment, which includes upcoming data protection laws and RBI guidelines, creates the conditions for the responsible implementation of AI. These are a few of these challenges and benefits: 10
CHALLENGES
Explainability and Transparency: The opacity of AI and machine learning (ML) algorithms can pose challenges, mainly when used in financial decision-making. Indian regulators emphasise the need for explainability and require banks to ensure that their AI models are transparent, understandable, and can be thoroughly examined for biases or errors. 8 Park, Jin & Thangam, Dhanabalan. (2019). What Makes Customers Repurchase Grocery Products from Online Stores in Korea. International Journal of E-Business Research. 15. 24-39. 10. 4018/IJEBR. 2019100102. 9 Malali, Anil B. , and S. Gopalakrishnan. "Application of artificial intelligence and its powered technologies in the Indian banking and financial industry: An overview. " IOSR Journal Of Humanities And Social Science 25. 4 (2020): 55-60.
10 Vidushi Marda, Artificial intelligence policy in India: a framework for engaging the limits of data-driven decision making, THEROYALSOCIETYPUBLISHING https://royalsocietypublishing. org/doi/full/10. 1098/rsta. 2018. 0087#d3e3
50 (Oct. 15, 2018).
Bias and Fairness: When AI models are trained using historical data, there is a risk of unintentionally introducing biases into the models. These biases may manifest in the form of discrimination when it comes to decisions related to financing and credit. It is crucial to address these ethical issues by implementing continuous monitoring and bias mitigation measures to ensure equitable outcomes for all demographic segments. • Consumer Consent and Control: It is essential to prioritise customers' consent and control over their information, especially considering the significant volume of personal data managed by AI systems. Regulatory frameworks must underscore the importance of clear and transparent communication with consumers about utilising their data.
Additionally, these frameworks should establish mechanisms that allow individuals to quickly reverse or modify their consent preferences. • Security and Cyber Threats: Given the increasing reliance on artificial intelligence in cybersecurity, it is imperative to develop comprehensive legal frameworks to combat potential cyber threats effectively. In particular, it is crucial for financial institutions to prioritise adherence to stringent cybersecurity standards and to be subject to legal accountability in the event of failure to protect customer data from unauthorised access.
BENEFITS
Proactive Administrative Compliance worldwide banking is heavily regulated to prevent financial crimes and ensure customer privacy. Banks must comply with rules regarding customer identification, money laundering prevention, privacy protection, and more. Compliance comes with high costs and responsibilities. Banks are turning to AI assistants to monitor transactions and maintain regulatory compliance to meet these obligations. AI has already significantly impacted fraud prevention and is helping banks stay ahead in the ever-evolving landscape of financial crime.
Better Dynamic for Loans and Credit Banks use AI-based frameworks to make more informed, secure, and profitable lending decisions. These AI systems analyse behaviours and patterns to assess creditworthiness, especially for individuals with limited credit history. However, there are challenges, such as bias in AI models and the difficulty in explaining decisions made by neural networks, which can impact customer experience. Financial institutions must address ethical considerations and regulatory requirements when implementing AI in lending processes. • Reducing Bank Working Expenses and Hazard The banking industry increasingly turns to robotic process automation (RPA) and AI technologies to streamline operations, reduce errors, and cut costs.
According to Accenture, by replacing manual processes with intelligent automation tools, banks can improve risk management, focus on higher-value tasks, and achieve significant operational savings. • Artificial Intelligence Collaborators for Investing Banks increasingly use AI for investment decisions, algorithmic trading, robo-advisors, user-friendly banking apps, and even podcasts. Although human expertise is still essential, AI is improving the industry and is expected to keep doing so in the future.
LEGALITY OF AI GLOBALLY AND IN INDIA
Global AI regulation is still primarily at the policy level. Position papers and policy statements on AI have been released by several governments, discussing how to weigh the advantages and risks. Implementing laws and regulations about AI presents challenging philosophical and practical issues. Many countries worldwide have started to struggle with similar conflicts in recent years. China is a front-runner and apparent leader in the race for artificial intelligence, if not the race for AI regulation. Different aspects of AI have been regulated differently in China. China established a set of rules about recommendation algorithms in 2021. Deep synthesis regulations were passed in 2022, while draft generative AI regulations were approved in 2023. The implementation of these laws is the responsibility of many entities. As a result,
China is establishing specific rules to serve as the foundation for a national AI law. This vertical strategy establishes separate rules to regulate various facets of artificial intelligence. 11 The European Commission-sponsored Artificial Intelligence High-Level Expert Group (AI HLEG) published its AI regulation strategy document, "Ethics Guidelines for Trustworthy AI," in April 2019. These recommendations guide any legal changes about AI use in the EU. They highlight several of the previously mentioned points, including the significance of user consent when implementing AI and transparency and privacy. Maintaining trustworthiness ultimately entails putting policies in place to ensure AI functions within the bounds of accountability, transparency, explainability, and fairness. 12 A fundamental set of guidelines for regulating AI is developing between the EC's AI HLEG.
And other nations. Among these essential components are the following:
1. Human judgment and supervision
2. robustness and security
3. Data governance and privacy
4. Fairness, diversity, and non-discrimination
5. Openness
6. Environmental and societal well-being
7. Responsibility
Article 22 of the General Data Protection Legislation (GDPR) of the European Union is the most explicit and robust example of AI legislation. GDPR, also known as the horizontal approach towards AI, is a comprehensive transnational regulatory framework that applies to all companies that deal with the data of EU residents, not just EU member states. Article 22, under the heading "Automated individual decision-making, including profiling," states the following:
11 Truby, Jon, Rafael Brown, and Andrew Dahdal. "Banking on AI: mandating a proactive approach to AI
regulation in the financial sector." Law and Financial Markets Review 14.2 (2020): 110-120.
12 Truby, Jon, Rafael Brown, and Andrew Dahdal. "Banking on AI: mandating a proactive approach to AI
regulation in the financial sector." Law and Financial Markets Review 14.2 (2020): 110-120.
1. The data subject shall have the right not to be subject to a decision based solely on
automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her.
2. Paragraph 1 shall not apply if the decision:
a. is necessary for entering into, or performance of, a contract between the data subject and a data controller; b. is authorised by Union or Member State law to which the controller is subject and which also lays down suitable measures to safeguard the data subject’s rights and freedoms and legitimate interests; or c. is based on the data subject’s explicit consent. Given the relative novelty of Article 22, concepts such as “legitimate interest” and “necessity for entering a contract” have yet to be fully developed. According to their challenges, the European Union's AI Act has divided artificial intelligence into three primary categories. The providers and users fall under distinct groups with corresponding obligations.
The three categories—social scoring, biometric or face identification, and cognitive behavioural manipulation of people—are unacceptable and seriously endanger human safety. This kind will not be permitted. The second category is high risk because it harms people's rights and presents a significant risk. The third type of AI is restricted risk, which calls for producers to disclose to consumers that they are engaging with AI and that the AI may utilise their data for content generation or manipulation. As a result, this is a more horizontal strategy, with a single piece of legislation covering every aspect of
AI.13
In contrast to these nations, India has very little regulation regarding AI. The Digital Personal Data Protection Act of 2023 is the lone exemption. Regarding data privacy abuse and violations, this Act governs a few facets of artificial intelligence. Nonetheless, it has its challenges. One of the main issues facing policymakers in India is how to govern AI in a way 13 Arora, Nalin, and Dhanishtha Arora. “Regulating AI in India: Navigating Global Approaches and Charting a Unique Path.” Jindal Society of International Law, 9 Apr. 2024
that sufficiently examines the current platforms and prevents them from becoming obsolete due to impending advancements in the field.14 There are issues with the EU's horizontal and China's vertical strategies. Business associations have criticised the EU's policy, claiming it hurts innovation. China's approach to information control is criticised for being overly strict. Another area for improvement with Chinese regulations is that some terms have been written so ambiguously that they may eventually become obsolete. This demonstrates the fact that strategy can only partially succeed. The components of both methods must be combined in a way that makes sense for the Indian situation, economy, culture, and legal system. Finding a suitable framework that strikes this balance without going overboard while protecting data privacy and fundamental rights is still challenging for India. 15
FUTURE TRENDS AND OPPORTUNITIES
The banking industry's early adopters of Al are now beginning to realise the real advantages of this technology. Future trends will surely pick up speed, so banks that have been sluggish in reacting will no longer be able to ignore Al. Nonetheless, banks need to address specific inherent issues with Al appropriately. Today's banking consumers expect their banks to provide the same excellent customer service they are accustomed to receiving from Big Techs. Consumers do not particularly like poor service, unpleasant encounters, or system failures. Consequently, a poorly thought out or executed Al investment could harm banks, possibly even working against them. Banks must take a thoughtful approach to Al that complements their broader IT strategy. Before an enterprise-wide implementation of Al solutions, known as an "Al Transform," banks must consider the concerns of all relevant parties.
According to the executives who participated in a poll across Indian banks, understanding Al is the second most important skill needed to execute Al, behind being flexible. It's interesting to note that tools that use Al algorithms, high-performance computing infrastructure, and Al- 14 Arora, Nalin, and Dhanishtha Arora. “Regulating AI in India: Navigating Global Approaches and Charting a Unique Path. ” Jindal Society of International Law, 9 Apr. 2024 15 Arora, Nalin, and Dhanishtha Arora. “Regulating AI in India: Navigating Global Approaches and Charting a Unique Path. ” Jindal Society of International Law, 9 Apr. 2024
related skills were regarded as significantly less critical. This suggests that these talents may be learned if one has a strong will to adapt and an awareness of the necessary tasks. 16 Banks must prepare the company for the change before conducting an Al Transformation. Several essential elements need to be taken into account: • Leadership and Governance Banks with a well-defined Al governance structure and an efficient Al leadership group are more likely to benefit fully from any Al implementation. Robust and established data procedures must be in place for banks to fully grasp Al's potential to change the game. The Al program needs a designated leader who can use Al as a significant organisational enabler by establishing the proper guidelines for data handling, governance, and gathering and an overarching vision.
According to Capgemini's research, companies with a dedicated Al lead outperformed those without one, where the different Al projects were operating as independent endeavours. This was the case for companies worldwide and across all business sectors. Furthermore, banks must invest in developing a group of Al experts with the appropriate skill sets and ensure enough pertinent data is available to train the Al systems. • Customer Impact Banks must consider how their consumers view Al and how the technology will affect them when developing their Al plans. Customers prefer interactions made possible by a blend of artificial and human intelligence, according to a recent Capgemini study on how customers view Al.
Furthermore, 83 per cent of financial services customers want to know that they are communicating with Al, whereas just 20 per cent of executives think that's what customers want to know. According to research, The following are the main advantages that Indian consumers cite for Al-enabled technologies:
1. Constant accessibility (67%)
AI
Revolution in Indian Banking.” Capgemini, https://www.capgemini.com/wpcontent/uploads/2019/02/IBABankTech2019AIRevolutioninBankingFinal.pd
2. Quicker handling of inquiries (56%)
3. Improved data security and privacy (52%)
In other words, Indian customers find it more credible when their Al-based interactions have human-like characteristics. 83% of customers desire a human-like cognition from their Al systems, 85% choose a human-like voice, and 82% want the capacity for their Al systems to recognise and react to human emotions. " However, while developing applications, most Indian banks prioritise data availability and return on investment over the needs of their customers. According to Capgemini's study, only 21% of businesses consider use cases that affect the customer experience, and 13% think about use cases that address well-known customer pain points. In the future, banks could wish to concentrate on implementing use cases that enhance the client experience.
According to Capgemini's research, 84% of Indian consumers shared their positive experiences, suggesting that having a positive Al-enabled experience will increase bank loyalty. Additionally, 78% of Indian consumers said they would be willing to spend more for a better experience, according to the study. • Employee Trust and Concerns Although the impact of Al on human occupations has been the subject of numerous discussions, research from Capgemini indicates that, worldwide, 82% of banks have had Al create new job opportunities in their businesses. In comparison, 72% have not experienced any job loss due to Al. " Al adoption in Indian banks has slowed down due to resistance to Al, which can also make institutions less receptive to change.
Previous studies have indicated that cultural concerns and stakeholders' resistance to change are the main obstacles to technology transformation rather than technological advancements. Banks must alleviate the worries of their staff members by maintaining an open communication channel. To help employees become more accustomed to the impending change and understand that people and computers will need to collaborate going forward, bank leadership should involve staff members at every stage of the process (but not at the expense of the other). Banks should also retrain their workforce to prepare employees for potential positions.
Conclusion India is currently at a critical juncture in the fast-evolving field of artificial intelligence. While countries and international organisations like China and the European Union have embraced different approaches to regulating AI, each facing its drawbacks and criticisms, India needs to devise a distinct strategy that fits its socio-cultural context, technological ambitions, and economic requirements, especially within the banking sector. Successful AI integration rapidly changes from a competitive edge to a crucial and immediate need in today's banking industry. The banks of the next generation must actively harness AI technology for successful AI integration. They should begin with use cases that can be quickly implemented and do not demand extensive technical expertise.
Banks must also carefully consider changes in their management structure, engage with all stakeholders, and provide the necessary tools and training to adapt to the changes brought about by AI.