An announcement about a document, not the document
Speaking at the inauguration of SEBI's pavilion at the 45th India International Trade Fair on Saturday, chairman Tuhin Kanta Pandey said the regulator will soon issue detailed guidelines for the responsible use of artificial intelligence in the capital markets. The guidelines, he said, would set guardrails covering disclosures, the accountability of regulated entities and a human-in-the-loop approach, so that critical decisions remain subject to human oversight.
Read the reports closely and that is the whole substance. There is no date attached to "soon." The reported remarks supplied no final text for firms to plan against. There is no list of which decisions count as critical, which is the distinction the entire human-oversight promise turns on: an AI system that drafts client emails and an AI system that places orders are both "AI in capital markets," and a rule that treats them the same would carry very different costs for each. There is also nothing on what happens to a firm that ignores the eventual guidelines.
None of that makes the direction wrong. SEBI is explicit that the aim is to balance adoption against investor protection, market integrity and financial stability. But direction is not regulation. The reported announcement identifies no newly issued obligation; it does not by itself establish the full set of existing requirements that already apply to AI use.
More than sixteen months since the consultation closed
The timeline is the part worth keeping. SEBI published its consultation paper on responsible AI and machine-learning use in securities markets on June 20, 2025. Public comments closed on July 11, 2025. The paper proposed five guiding principles: model governance, investor protection through disclosure, a testing framework, fairness and bias controls, and data privacy and cyber security measures. It also floated a lighter touch for AI uses that do not directly affect customers, against full expectations for client-facing uses such as algorithmic trading, portfolio management and advisory tools.
That was a workable starting point, and it has now been the latest word for more than 16 months. Saturday's remarks confirm the project is alive and restate its three headline themes. They do not explain what the consultation responses changed or which principles survived. A reader who followed the 2025 consultation learns one new fact from this announcement: there is still no finished document.
Pandey also pointed to work SEBI is doing on its own side of the table, including supervisory tools such as its "Sudarshan" system and a structured cyber-incident response mechanism introduced in May 2026, aligned with the Financial Stability Board's FIRE reporting template. That is a separate claim from regulating the market's AI, and it should not be counted as progress on the guidelines themselves. A regulator using AI to watch the market and a regulator publishing enforceable rules for the market's AI are different deliverables.
What to watch for in the actual text
When the guidelines do appear, ALFW's evaluation checklist starts with four details that will decide whether they bite. First, the definition of a critical decision, and whether human oversight means a person approves each action or a person is accountable after the fact. Those are very different costs. Second, whether accountability stays with the regulated entity when the model comes from an outside vendor. Third, what disclosure means in practice: telling clients that AI is used somewhere in the firm, or telling them when it shaped their specific advice or trade. Fourth, a start date with a transition period, rather than an expectation that applies on publication.
Until those four points are in print, treat "SEBI AI guidelines" headlines as a status update on a 2025 consultation, not as a new obligation. Firms operating in Indian markets can document their AI use against the five consultation principles, which remain proposals, not binding standards. Investors reading that a firm is "SEBI AI compliant" should ask compliant with what: as of November 15, 2026, the reported announcement identifies no final SEBI AI guideline to comply with.
Sources
- [1] The Hindu BusinessLine, "SEBI to issue AI norms for capital markets, safeguard investor interests: Pandey," November 15, 2026Read source
- [2] SEBI, "Consultation Paper on guidelines for responsible usage of AI/ML in Indian Securities Markets," June 20, 2025Read source
- [3] Storyboard18, "SEBI to issue AI guidelines for capital markets, prioritise investor protection," November 15, 2026Read source
- [4] BFSI Network via Elets, "Sebi to issue AI guardrails for capital markets with human oversight mandate," November 15, 2026Read source