Technology in the post-industrialization era was deemed the next big revolution in easing our day-to-day activities and it has not disappointed us. The meteoric rise of Artificial Intelligence (AI) in multidisciplinary fields demonstrates the power of adaptability that AI retains. The advancement of AI is based on big data which is processed efficiently and managed due to increased computing speeds and constantly improving algorithms.
Machine Learning (ML) has been an epicentre in the storm of AI that has taken over the world. ML which acts as a sub-branch to AI, holds the knowledge to self-educate the AI personalized program to self-educated based on the interaction and experience. AI for instance has been integrated into various functionalities like services, decision-making, business management, data-based predictability support, assistance, etc. which is scattered into various sectors varying from healthcare, manufacturing, and education to finance.
The buzzing terminology “FinTech” has been constantly ringing the news as financial institutions integrated advanced tech into them. Fintech encompasses all financial services and products including but not limited to lending, investing in stocks, payments, insurance, and financial information generated or published through these mediums.
What sparks SEBI’s interest in AI and ML?
The Securities and Exchange Board of India (SEBI) has been constantly concerned with the advancement of AI & ML in the investing arena. The consultation paper released by SEBI on Nov 13th, 2024[1] proposes amendments that confer SEBI the power to call out any risks of financial lapses. AI as often is used to study the stock and its financials, past trends, growth of the company, and edge over competitors, etc. These learnings are used to predict the market which in itself is unpredictable in the most efficient way leading to potential gains and minimizing losses. AI/ML can be used in combination to generate predictive modelling which can be used to analyze and anticipate the potential output instantaneously in real-time based on the data fed into it. In the finance sector, the players have been utilizing the services of AI/ML to effectively increase functionalities and efficiency like automating customer services, filtering potential investment stocks through AI, offering financial advice through AI, assisting with decision-making, producing analysis, etc.
However, AI is more susceptible to malicious attacks, where malicious actors intentionally manipulate financial data to deceive AI and harm potential investors(s). Particular AI leans heavily on large amounts of sensitive investor data. The more the data, the output will be more accurate but with a potential downside of sharing the investor’s data. Certain AI/ML possess the functionality of self-learning, free from human supervision, which could trigger engaging in market manipulation, conflicts of interest, or other unlawful activity. SEBI’s consultation paper highlights AI/ML implications that could potentially lead to ethical dilemmas, security vulnerabilities, and potential misuse in the financial sector affecting investors and stakeholders. To address these concerns and enhance the protection shield around investors, it has proposed the below amendments. SEBI’s concern revolves around the lack of transparency surrounding the advanced AI algorithms and their implications.
Highlights of the proposed amendments
SEBI had previously issued circulars concerning the increased usage of AI/ML systems by mutual funds[2] and market intermediaries[3] In their services to investors or consumer products in 2019. Now in 2024, SEBI seeks to allocate accountability to the market intermediaries, institutions, and regulated entities to assure investor protection.
- Obligations for compliance: The financial frauds in India are often linked to government failures citing the dearth of supervision, regulations, compliance, oversight, and lack of accountability. SEBI through the proposed amendments wants the Regulated Entities (RE) to be held ‘accountable’ for any irregular consequence generated or derived resulting from the use of AI/ML that directly affects the investor.
- Liability for AI outcomes: People and businesses are leaning on AI to make financial decisions based on their choices. Hence, SEBI proposes that all RE should make sure that all AI/ML-related tech, software, and tools are compliant with the relevant laws and jurisdiction. In cases of non-compliance, the RE will bear complete responsibility for all consequences arising from the use of AI, irrespective of the extent of its usage.
- Accountability for breach/leak of investor data: AI/ML technology models feed on the type, size, and quality of data provided to them to get the desired outcome. The proposed amendments highlight SEBI being distressed about the critical, financial, and personal data of an investor. Therefore, SEBI suggests regulating the privacy, security, and integrity of the investor’s and stakeholders’ data, especially the data held in a fiduciary capacity by making the RE accountable for any breach.
Concerns surrounding SEBI’s amendments
In response to SEBI’s consultation paper, the Asia Securities Industry and Financial Markets Association (ASIFMA) presented concerns.[4] That centred on the issue of accountability. ASIFMA argued that the proposed regulations should adopt a flexible approach, as the use of AI varies significantly across entities and institutions. ASIFMA endorsed the idea of holding financial institutions accountable for their misconduct, but it emphasized a shared responsibility framework for instances where AI produces fair and accurate outputs, but investors still make flawed judgments. In such cases, ASIFMA underlined that it would be unjust for institutions to be held liable for outcomes beyond their control.
There’s also a concern that RE might use the notion of “beyond their control” as a defence against liability. AI/ML systems, though autonomous in operation, are ultimately designed, programmed, and maintained by humans. Allowing institutions to deflect responsibility could erode trust in the financial system, as investors might feel unprotected against the traps of complex AI/ML technologies they don’t fully understand.
The approach to regulation, holding RE completely responsible without accounting for I/ML usage extent, seems overly stringent. AI/ML technologies, when appropriately utilized, offer significant advantages, such as preventing financial fraud through predictive algorithms, assessing credit risk, analyzing large financial reports for irregularities, errorless regulatory information, lowering human errors, and managing customer interactions 24/7. These advancements not only improve operational efficiency but also enhance investor protection. However, the regulatory inconsistency concerns raised by ASIFMA question the inability of the scarcity of such laws and orders regulating AI/ML in India.
Conclusion
SEBI’s open dialogue system for governance related to technology in the finance sector is truly appreciable with the stock market regulator inviting comments for feedback sheds a positive light. A balanced framework in regulating the accountability of AI/ML along with human oversight coupled with transparency could benefit in slicing the accountability further resulting in seamless transactions, boosted efficiency, fraud-less activities, and errorless compliance.
Author(s) Name: Pranav Wadhawankar
References
[1] Securities and Exchange Board of India, ‘Proposed Amendments with Respect to Assigning Responsibility for the Use of Artificial Intelligence Tools by Market Infrastructure Institutions, Registered Intermediaries and Other Persons Regulated by SEBI’ (2024) https://www.sebi.gov.in/reports-and-statistics/reports/nov-2024/proposed-amendments-with-respect-to-assigning-responsibility-for-the-use-of-artificial-intelligence-tools-by-market-infrastructure-institutions-registered-intermediaries-and-other-persons-regu accessed 4 February 2025.
[2] Securities and Exchange Board of India, ‘Reporting for Artificial Intelligence (AI) and Machine Learning (ML) Applications and Systems Offered and Used by Mutual Funds’ (2019) https://www.sebi.gov.in/legal/circulars/may-2019/reporting-for-artificial-intelligence-ai-and-machine-learning-ml-applications-and-systems-offered-and-used-by-mutual-funds_42932.html accessed 4 February 2025.
[3] Securities and Exchange Board of India, ‘Reporting for Artificial Intelligence (AI) and Machine Learning (ML) Applications and Systems Offered and Used by Market Intermediaries’ (2019) https://www.sebi.gov.in/legal/circulars/jan-2019/reporting-for-artificial-intelligence-ai-and-machine-learning-ml-applications-and-systems-offered-and-used-by-market-intermediaries_41546.html accessed 4 February 2025.
[4] Business Standard, ‘FPI Lobby Opposes SEBI’s Proposed Norms for Regulating AI and ML’ (2025) https://www.business-standard.com/markets/news/fpi-lobby-opposes-sebi-s-proposed-norms-for-regulating-ai-and-ml-124120600981_1.html accessed 4 February 2025.
