SEBI issues guidelines on AI use in capital markets

SEBI is working on a comprehensive framework to govern the use of artificial intelligence in the capital markets ecosystem, SEBI chair Tuhin Kanta Pandey said at the ET NOW Markets Summit 2026, per The Economic Times. Pandey was quoted saying, "AI will be an important part of our regulatory agenda. AI can improve surveillance, risk assessment, fraud detection, and investor servicing." He also warned that AI "brings risks relating to opacity, bias, data protection, cybersecurity, and accountability." The Economic Times reports that SEBI "will issue detailed guidelines on the responsible use of AI in capital markets." Pandey said IOSCO's AI supervisory toolkit would be suitably integrated into SEBI's AI strategy for regulated entities, and that an expert panel will inform a five- and 10-year roadmap for market infrastructure institutions.

Industry-pattern observations: Financial regulators globally have focused guidance on model explainability, data governance, and third-party vendor risk for systems used in surveillance and trading. Practitioners deploying ML in trading surveillance, fraud detection, or investor servicing commonly need to document model lineage, bias testing, and incident response capabilities to satisfy such frameworks. Integration of an international toolkit such as IOSCO's typically points to emphasis on supervisory checkpoints rather than permissive innovation alone, based on prior cross-jurisdictional regulatory practice.

Editorial analysis: National regulators issuing AI rules for capital markets tends to raise compliance and operational requirements for exchanges, broker-dealers, market infrastructure providers, and fintech vendors. For data teams and ML engineers in these firms, regulatory guidance usually increases demand for explainability tooling, robust data provenance, and formal model risk-management processes. Past regulatory efforts in financial services prioritized controls around data privacy, model validation, and vendor governance, which is likely to shape implementation choices where AI is embedded in decision-making.

• •Whether SEBI publishes the detailed guidelines and their timeline, as referenced by Pandey in The Economic Times.
• •The specific obligations that mirror IOSCO's AI supervisory toolkit versus locally tailored requirements for Indian market infrastructure institutions.
• •How exchanges and regulated entities translate guidance into technical controls around model monitoring, retraining cadence, and incident reporting.