The emergence of AI-powered trading agents that can independently modify their strategies as they learn could pose fresh regulatory challenges for capital markets, Securities and Exchange Board of India (SEBI) Whole-Time Member KVR Murty said on Thursday, signalling the need for a new framework to ensure accountability and traceability of such systems.
Unlike conventional algorithms that operate on pre-defined rules, AI agents are given an objective and can change how they trade as they learn. This raises a fundamental question for regulators: how do they approve a trading strategy that can rewrite itself?
Addressing a World Federation of Exchanges (WFE) session in Mumbai on the next frontier for capital markets, Murty also flagged the risk of a large number of investors using AI agents built on the same models and datasets, potentially triggering simultaneous trading decisions and market volatility.
"An AI agent is given a goal, not a fixed set of rules, and it can change how it trades as it learns. So how do you approve a strategy that rewrites itself? And what happens when a million investors run agents built on the same few models and the same data, and all head for the exit at once? We don't have every answer yet, and I doubt any regulator in this room does," he said.
"But one principle will not change: every agent, however intelligent, must be traceable, and someone must answer for what it does," he added.