The Securities and Exchange Board of India (SEBI) is preparing guidelines on AI use in capital markets that would cover disclosures and accountability for related entities, whole-time member KVR Murty said at a World Federation of Exchanges event in Mumbai on 19 November, Maine Mirror reported.
Key points
- SEBI is preparing guidelines covering disclosures and accountability for entities using AI in capital markets.
- Murty questioned how a regulator could approve a trading strategy that changes as an agent learns.
- He called for agents to be traceable and for someone to answer for their actions.
Murty questions approval of changing strategies
An AI trading agent can alter its approach as it learns rather than follow a fixed set of instructions, Murty said. He asked how SEBI could approve a strategy capable of rewriting itself, Moneycontrol reported. Murty also raised the prospect of investors using agents built on similar models and data leaving the market at the same time.
Murty said every agent must be traceable and someone must answer for its actions. Where an algorithm’s decision affects a customer or the market, the organisation responsible must be prepared to give its reasons, he said.
SEBI’s proposed disclosures and human oversight
SEBI chairman Tuhin Kanta Pandey had discussed the planned guidelines at the 45th India International Trade Fair on 14 November. The proposal would address disclosures by market participants, accountability for regulated entities and human intervention in critical decisions, Storyboard18 reported. SEBI released a consultation paper on responsible AI and machine-learning use in securities markets in June 2025. AI Affairs previously reported Pandey’s plans for guidelines with human oversight.
Murty also identified market surveillance, compliance and fraud detection as potential uses for AI, while saying governance frameworks remain incomplete.