• EXECUTIVE SUMMARY

Artificial intelligence (AI) is rapidly reshaping financial markets by enhancing pricing, execution, surveillance, documentation and risk management. It enables institutions to process large volumes of structured and unstructured data, identify market patterns, automate trading strategies and strengthen real-time monitoring. At the same time, AI introduces material legal and regulatory risks, particularly because models may be opaque, adaptive, vendor-dependent and difficult to supervise. For Indian financial markets, the relevant question is therefore no longer whether regulated entities should use AI, but how they should govern its use. The emerging regulatory position is clear: institutions remain accountable for AI-enabled decisions, even where the model is developed, hosted or operated by a third party.

• AI IN FINANCIAL MARKETS

Financial markets have long relied on quantitative models, professional judgment and regulatory safeguards to support pricing, valuation, execution and risk management. AI, particularly machine learning and generative AI, is accelerating this evolution by making market analysis more data-driven, adaptive and responsive. In pricing and valuation, AI can draw on wider datasets, identify complex relationships and respond more quickly to changing market conditions, thereby supporting price discovery and risk analytics, especially where market information is fragmented or fast-moving.

AI is also transforming trading behaviour. Algorithmic and automated trading systems increasingly use predictive analytics to identify patterns, generate strategies and execute trades with varying degrees of human intervention. These systems can improve liquidity, reduce spreads and lower transaction costs. However, they may also amplify volatility, encourage algorithmic herding and accelerate market dislocations during periods of stress. The regulatory challenge is therefore to preserve efficiency and innovation while safeguarding market stability, fairness and systemic resilience.

Risk management is another significant area of transformation. Traditional frameworks rely heavily on historical data, static stress scenarios and periodic review. AI enables more dynamic and forward-looking monitoring by identifying anomalies, vulnerabilities and emerging patterns in near real-time. It can support exposure monitoring, liquidity assessment, collateral optimisation, counterparty risk evaluation and stress testing. However, reliance on opaque algorithms may create model risk, particularly where errors in alerts, valuations, risk limits or stress thresholds could have broader consequences across interconnected markets.

• REGULATORY RESPONSE AND ACCOUNTABILITY

India’s regulatory approach is moving from a largely technology-neutral framework towards more express governance of AI, machine learning (ML) and model risk. The Reserve Bank of India’s (RBI) draft[1] guidance on model risk management for regulated entities is particularly relevant because pricing, valuation, collateral optimisation, exposure measurement, suitability assessments and automated risk decisions may all be model-driven. The guidance points towards lifecycle-based model governance, including board-approved frameworks, clear ownership, model inventories, risk-based classification, independent validation, periodic review, human oversight and controls for AI and ML models, including customer-facing and generative AI use cases.

The practical implication is that AI cannot be treated merely as a technology tool. Where an AI model influences a material financial decision, the regulated entity must be able to explain, validate and control it. Outsourcing the model does not outsource regulatory responsibility. AI outputs used for material decisions should therefore be supported by explainability, audit trails, independent challenge, escalation mechanisms and the ability to restrict, suspend or decommission unreliable models.

Securities and Exchange Board of India’s (SEBI) framework[2] for algorithmic trading and AI/ML reporting remains relevant for AI-enabled trading and market activity. It emphasises exchange approval or registration of algorithms, risk checks, audit trails, traceability of orders, reporting of AI/ML systems and safeguards against manipulation or disorderly trading. International Financial Services Centres Authority’s (IFSCA) role is also important in the Gujarat International Finance Tec-City International Financial Services Centre (GIFT City IFSC) ecosystem, where AI usage may raise additional questions around cross-border data flows, outsourcing, supervisory access and the integration of global algorithmic systems within the Indian regulatory perimeter.

Internationally, regulators appear to be converging on the same broad position. The EU has adopted a risk-based framework under the EU AI Act[3], while the UK continues to rely on a principles-based approach through existing financial regulatory toolkits. Singapore[4] and Hong Kong[5] have similarly focused on responsible AI, data ethics, model governance, outsourcing controls, cybersecurity, explainability and customer protection. Across jurisdictions, the common concern is not only whether an AI model works, but whether it can be understood, tested and supervised.

• LEGAL AND SYSTEMIC RISKS

The growing use of AI introduces several legal and systemic risks, including model risk, explainability risk, market conduct risk, outsourcing and vendor risk, cybersecurity risk, data protection risk and concentration risk where multiple institutions rely on similar models or service providers. A key legal question is who bears responsibility when an AI system produces an incorrect valuation, executes an erroneous strategy, triggers a risk action, denies or recommends a product, or contributes to disorderly market behaviour. The emerging answer is that responsibility cannot be shifted entirely to the model, vendor or technology team. The regulated entity deploying or relying on the AI system remains primarily responsible for its governance, use and consequences.

Responsibility should therefore be allocated clearly within the institution. The board and senior management should approve the governance framework and risk appetite. Business teams should ensure that AI tools are used within approved limits. Risk, compliance and validation teams should test, monitor and challenge model outputs. Technology and data teams should ensure system integrity, cybersecurity and data quality. Vendors may assume contractual obligations, but they do not dilute the regulated entity’s accountability.

• CONCLUSION

AI is likely to become an integral part of financial markets, influencing automated trading, risk systems, alternative data use, market surveillance and supervisory scrutiny of algorithmic behaviour and model governance. Its benefits are substantial, but they must be balanced with robust controls around transparency, explainability, validation, auditability, human oversight, vendor management and systemic resilience.

The key takeaway is that AI may improve the speed, scale and sophistication of financial markets, but it does not reduce the responsibility of regulated entities. Institutions that deploy, rely on or benefit from AI systems must retain effective control over model approval, validation, monitoring, escalation, suspension and remediation. The future of financial markets will therefore depend on balancing innovation and efficiency with institutional accountability and regulatory oversight.

[1]  RBI Press Release dated 24 June 2026, Draft Guidance on Regulatory Principles for Model Risk Management.

[2]  SEBI: Reporting for Artificial Intelligence (AI) and Machine Learning (ML) applications and systems offered and used by market intermediaries & SEBI: Consultation Paper on guidelines for responsible usage of AI/ML In Indian Securities Markets.

[3]  EU AI Act (Regulation (EU) 2024/1689).

[4]  Monetary Authority of Singapore, Principles to Promote Fairness, Ethics, Accountability and Transparency (FEAT) in the Use of AI and Data Analytics in Singapore’s Financial Sector (2018).

[5] Hong Kong Monetary Authority, High-level Principles on Artificial Intelligence (2024).

Authors:

Disclaimer: 

This article is intended for informational purposes only and does not constitute a legal opinion or advice. Readers are requested to seek formal legal advice prior to acting upon any of the information provided herein. This article is not intended to address the circumstances of any particular individual or corporate body. There can be no assurance that the judicial / quasi-judicial authorities may not take a position contrary to the views mentioned herein