SEBI signals a forthcoming tiered framework for responsible AI use in securities markets, with guidelines that require clear accountability, data controls, human oversight, and a kill switch.

Regulated entities will be accountable for the AI/ML systems they use, including protecting investor data privacy, security and integrity, and responsibility for outputs, whether systems are developed in-house or sourced from third parties.

Regulations aim to balance innovation with investor protection and market integrity as AI tools proliferate in trading and financial services.

Capital-market reforms seek broader participation and financing channels, including potential foreign investment in portfolios, digital onboarding for non-residents, expanded FPI access to non-agricultural derivatives, tokenisation of corporate bonds, and greater REIT/InvIT exposure to projects under construction.

Growth indicators cited include AIF investments around Rs 7 lakh crore by late July 2026, market capitalization about 132% of GDP, roughly 14.9 crore unique investors, mutual fund assets near Rs 86 lakh crore, and SIPs making up over 20% of industry assets.

The remarks were delivered at the 23rd FICCI CAPAM 2026 event.

SEBI is expanding API-based connectivity and interoperability among market participants and pursuing IT resilience for market infrastructure, including an IT resilience index.

Pandey advocates deeper financialisation, broader participation, and new investment frontiers like infrastructure, technology, energy transition, data centers, and AI-driven opportunities.

Guidelines aim to balance innovation with investor protection as India’s capital markets expand, supported by hundreds of IPOs and large mutual fund assets.

Current market activity is strong, with significant IPO and corporate-bond issuance, underscoring the need for robust investor trust and governance.

SEBI is considering an IT resilience index to objectively assess the resilience of market infrastructure technology systems.

SEBI already uses AI tools like SUDARSAN and R(AI)DAR for surveillance and is extending governance of AI to external vendors and products, with cyber-resilience initiatives like the Cyber Suraksha Portal.