The Securities and Exchange Board of India (SEBI) is preparing to issue guidelines for the responsible use of artificial intelligence and machine learning in India’s capital markets, according to SEBI Chairman Tuhin Kanta Pandey. The proposed framework will take a tiered approach, with requirements around accountability, governance and investor protection as financial institutions increasingly deploy AI.
A major feature will be mandatory human oversight and “kill-switch” mechanisms. Market participants using AI systems would need controls that allow humans to intervene or rapidly disable an AI system when it behaves unexpectedly. The framework will also include data controls, reflecting concerns about the quality, security and appropriate use of information feeding AI systems.
SEBI sees significant potential for AI in market surveillance, risk assessment, fraud detection and investor servicing, but Pandey also highlighted risks involving opacity, bias, cybersecurity, data protection and accountability. Importantly, regulated entities would remain responsible for AI and ML systems they deploy, including tools developed by third-party providers. This prevents firms from shifting responsibility to an AI vendor when an automated system causes a problem.
The announcement builds on SEBI's earlier work: the regulator published a 2025 consultation paper on responsible AI/ML use in Indian securities markets, and has also been developing AI-based regulatory and surveillance capabilities. The broader direction is clear: India is not trying to stop AI adoption in financial markets, but is moving toward a model where AI can automate and assist, while humans retain ultimate control and accountability.
