The Securities and Exchange Board of India (SEBI) has initiated a significant regulatory process by issuing a consultation paper on 13 November 2024, targeting the governance of artificial intelligence (AI) and machine learning (ML) technologies in financial markets. This proposal emerges from a systematic approach SEBI has been developing since 2019 during which the regulatory body has been meticulously tracking AI and ML applications across various market participants including stockbrokers, depositories, clearing corporations, and mutual funds.
The proposed amendments are fundamentally driven by several critical concerns surrounding the implementation of AI technologies. Foremost among these are challenges related to data quality and integrity, where even minor alterations in data sets can dramatically transform AI system outputs. Additionally, the regulatory body is deeply concerned about the inherent lack of transparency in advanced AI algorithms, often referred to as 'black box' systems, which make decisions without clear explanatory mechanisms.
SEBI's draft amendments predominantly focus on establishing comprehensive accountability for regulated entities (REs). The proposed framework mandates strict compliance with existing legal standards, robust protection of data privacy, and complete operational liability for outcomes generated by AI and ML technologies. This approach signifies a proactive stance in ensuring technological innovations do not compromise investor interests or market integrity.
The Asia Securities Industry & Financial Markets Association (ASIFMA), representing foreign portfolio investors, has mounted a robust challenge to these proposed norms. The industry lobby argues that SEBI's approach represents a potentially problematic one-size-fits-all regulatory strategy that could inadvertently stifle technological innovation. ASIFMA advocates for a more nuanced, shared responsibility framework where liability is distributed across different stages of the AI value chain.
Central to ASIFMA counterargument is the concern over holding financial institutions accountable for client decisions made using AI-generated outputs. The organisation argues that it would be unreasonable to impose liability when an AI tool provides accurate information, but a client subsequently makes an independent decision. This perspective challenges SEBI's current proposed language which seems to suggest comprehensive institutional responsibility.
The industry body has proposed several constructive alternatives to SEBI's approach. These include adopting the OECD's (Organisation for Economic Co-operation and Development's) definition of AI systems, implementing a more granular responsibility model, and ensuring that the proposed regulations align with international regulatory practices. ASIFMA is particularly concerned that the current proposal might represent a significant regulatory overreach that diverges from global standards.
For financial institutions, these developments signal a critical need for strategic recalibration. Experts recommend exercising heightened caution in AI technology development, implementing robust human oversight mechanisms, and conducting thorough evaluations of third-party AI solutions. The proposed regulations underscore the delicate balance between technological innovation and regulatory protection.
The ongoing consultation process represents a pivotal moment in India's approach to technological governance in financial markets. By engaging in this dialogue, SEBI demonstrates a commitment to creating a regulatory framework that can adapt to rapidly evolving technological landscapes while maintaining core principles of investor protection and market integrity.
As the consultation continues, stakeholders across the financial technology ecosystem will be closely monitoring the potential refinements to these proposed amendments. The ultimate goal remains creating a regulatory environment that fosters innovation while providing comprehensive safeguards for investors and market participants.
