SEBI's AI/ML Framework: Balancing Innovation, Compliance, And Market Integrity

Introduction

Artificial Intelligence (AI) and Machine Learning (ML) are changing the India securities market from how things operate, influencing various facets including trading algorithms, portfolio analysis, compliance checks, and investor services. These improvements have enhanced decision-making processes, elevated market efficiency, and increased liquidity. However, they also pose some complex regulatory challenges like algorithmic transparency, systemic vulnerabilities, and the risk of market manipulation highlighting the urgent need for regulatory frameworks that is at par with the technological advancements.

In response, the Securities and Exchange Board of India (SEBI) introduced an innovative approach by establishing a thorough AI and ML reporting framework in 2019. The purpose was not to limit technological innovation, but to assure that the use of AI and ML by intermediaries is verifiable, and in line with the principles of market integrity. Over time, SEBI and the stock exchanges have systematically enhanced and diversified this framework, leading to the formation of an integrated, cross-exchange oversight system in 2025, a system that reconciles innovation, accountability, and investor protection amidst India's fast-paced digital market dynamics.

From Regulatory Vision to Digital Oversight: SEBI's AI & ML Supervision Journey

SEBI's directive in 2019 signified a pivotal moment at the convergence of technology and financial regulation. Via Circular No. SEBI/HO/MIRSD/DOS2/CIR/P/2019/10 dated January 4, 2019, the regulatory body transitioned from mere observation to proactive governance of AI and ML applications within India's securities framework. It required all recognized exchanges including NSE, BSE, MSE, MCX, and NCDEX to establish a cohesive supervisory system capable of overseeing how intermediaries utilize AI and ML tools in trading, risk assessment, and investor engagement. This initiative was a crucial component of SEBI's larger Enhanced Supervision Programme, aimed at promoting accountability, operational resilience, and technological integrity among market participants.

The Metropolitan Stock Exchange (MSE) was one of the first to accomplish the vision, issuing comprehensive circulars in February and April 2019 that translated policy intentions into actionable measures. The newly established mechanism mandated that intermediaries report their use of AI and ML every six months, ensuring that innovation remained in sync with regulatory oversight. Additionally, even firms lacking AI or ML systems were required to submit NIL reports, reinforcing the principle of universal compliance. By April 2019, the framework advanced into a fully digital procedure through the Exchange's Enhanced Supervision Portal, allowing for streamlined submission, validation, and audit trails signalling the dawn of a technology-driven regulatory landscape where compliance itself became data-driven.

Expansion and Harmonisation by the Exchanges

In December 2023, the MSE notably enhanced its reporting framework with the issuance of Circular No. MSE/MEM/14585/2023. The updated reporting template was synchronized with SEBI's fortified supervisory directives, while the definition of "AI/ML systems" was broadened to include predictive, analytical, and pattern-recognition tools utilized across front-mid- and back-office functions. This change acknowledged that AI/ML applications have now expanded beyond trading to also embrace decision-making, monitoring, and operational oversight.

The revision further emphasized a heightened commitment to transparency and traceability. Members are now obligated to uphold thorough internal approval processes, validation records, and system-level documentation for each AI/ML application submitted. This modification signified a distinct shift from simple compliance reporting to ongoing governance, fostering a culture of accountability within the internal compliance frameworks of intermediaries. This made compliance process easier specially for those operating across multiple exchanges. By 2025, NSE implemented a consolidated reporting framework designed to enhance the efficiency of AI/ML disclosures across Indian exchanges. Within this structure, trading members registered with the NSE were permitted to submit a singular AI/ML reporting document, which the Exchange would thereafter disseminate to other recognized exchanges, including BSE, MSE, MCX, and NCDEX. The MSE formally aligned to this model through Circular No. MSE/MEM/17921/2025, thereby integrating earlier directives and explicitly delineating procedural responsibilities for members engaged with multiple exchanges.

This streamlined approach significantly minimizes unnecessary filings and eases the compliance requirements for entities operating across multiple platforms. However, members who are not registered with the NSE are required to continue submitting directly to the MSE via its Enhanced Supervision Portal. The circular further reiterates that AI/ML disclosures remain mandatory for all algorithmic trading members including NIL reports.

Compliance Obligation

Under the consolidated AI/ML governance framework, every trading member utilizing algorithmic software is obligated to disclose its application of AI or ML systems for each semi-annual reporting cycle regardless of whether these systems are developed in-house or acquired from external providers. Even those members who do not implement any such technology must submit a formal NIL declaration, verifying non-usage during the reporting timeframe. All submissions are conducted online via the Enhanced Supervision Portal. Members have the option to either choose the NIL declaration or fill out the detailed data forms that specify each AI/ML system's purpose, operational functions, risk management measures, and vendor information. Furthermore, firms are required to keep thorough internal records that support their submissions encompassing design approvals, audit trails, risk-control frameworks, and vendor contracts which must be readily accessible for inspection by SEBI or the Exchange when necessary. This framework highlights that compliance involves more than just the accuracy of filings. It imposes a dual obligation on intermediaries ensuring adherence to procedures while also demonstrating ongoing oversight and control over the technologies they implement thus harmonizing innovation with governance and accountability.

Benefit for Traders

1. Enhanced market transparency and fairness: The obligatory disclosure of AI/ML systems guarantees a clear trading landscape. Traders gain from more equitable market practices, as all algorithmic instruments undergo validation and compliance assessments, thereby diminishing the chances of manipulation or unfair competitive edges.
2. Greater system reliability and reduced operational risk: The framework's stipulation for validation records and audit trails bolsters system integrity. Traders face fewer algorithmic malfunctions, enjoy increased reliability in execution, and experience a lower risk of encountering rogue or faulty AI systems.
3. Faster and more reliable market execution: The digitization and centralization of submissions via the Enhanced Supervision Portal heighten the efficiency of exchange supervision. Traders benefit from more seamless operations and diminished execution lags owing to improved regulatory oversight.
4. Protection and assurance for investors: A transparent and accountable AI/ML landscape fosters investor confidence. This results in enhanced liquidity and tighter spreads, indirectly aiding active traders by amplifying market depth and engagement.

How registered firms can benefit from this?

1. Reduced compliance burden through cross-exchange filing: The 2025 harmonised framework enables NSE-registered members to submit a single AI/ML disclosure that is automatically shared with other exchanges (BSE, MSE, MCX, NCDEX). This greatly minimizes redundant filings and reduces compliance expenses.
2. Improved governance and risk management: Companies are required to uphold thorough internal documentation, which includes approval trails and validation records. This enhances governance, risk evaluation, and operational resilience within the organizations.
3. Regulatory certainty and predictability: With distinctly standardised reporting obligations, companies achieve clarity and predictability in their compliance efforts. This facilitates more confident planning for the implementation and upgrades of AI/ML systems.
4. Data driven decision making: The organised reporting process produces essential operational and compliance data. Companies can utilise this information for internal analytics, performance assessment, and audit readiness.
5. Reputation and investor confidence: Adhering to SEBI's AI/ML governance improves a company's credibility and standing. Transparent operations demonstrate technological and ethical sophistication, thereby strengthening the confidence of investors and stakeholders.

Conclusion

The unified AI/ML reporting framework signifies an advanced regulatory achievement which intends to reinforce transparency and systemic stability within Indian securities market. By establishing a centralized repository for AI/ML systems, it facilitates oversight of intricate automated strategies, guarantees auditability during post-event evaluations, and strengthens accountability by obliging intermediaries to validate and oversee their tools whether they are created in-house or procured from external sources.

Since its launch in 2019, SEBI's AI/ML supervision framework has evolved into a cohesive model that adeptly balances technological advancement with regulatory oversight. Through ongoing disclosure obligations, enhanced documentation standards, and improved cross-exchange collaboration, SEBI and the exchanges have elevated India's capital markets to align with global benchmarks that emphasize explainability, ethics, and responsible automation. The overarching message is unmistakable: while innovation in algorithms is embraced, it must proceed alongside transparency, governance, and accountability.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.