AI Tools in Investment Advisory — What SEBI Requires You to Disclose to Clients

Whether it's a portfolio-screening model, an AI-assisted research summariser, or a client-facing chatbot, SEBI's regulatory framework doesn't wait for AI adoption to become universal before regulating it. For Investment Advisers, the disclosure obligation is already live, specific, and named — unlike the more general liability standard sitting behind it.

The Specific Rule — Regulation 15(14)

Amendments to the SEBI (Investment Advisers) Regulations, effective 16 December 2024, inserted Regulation 15(14), requiring IAs to disclose the extent of their AI tool usage to clients at the time of entering into an advisory agreement. Existing clients had to be brought into compliance with this disclosure shortly after the amendment took effect. This is a client-facing, agreement-level disclosure obligation — not a general marketing statement, but something that needs to sit inside the actual client onboarding and agreement documentation.

The Broader Liability Standard Sitting Behind It — Regulation 16C

Separately, the SEBI (Intermediaries) (Amendment) Regulations, 2025 inserted Regulation 16C into the Intermediaries Regulations, 2008 — a considerably broader provision applying to any SEBI-regulated person using AI/ML tools or techniques, whether built in-house or procured from a vendor. Under this provision, the regulated entity is solely responsible for the privacy, security, and integrity of investor data processed through the tool, and for the outputs the tool generates — a vendor's own terms of service or limitation-of-liability clause doesn't shift this accountability away from the IA.

What This Means Practically for an IA Using AI Anywhere in the Client Workflow

If AI is used for portfolio analysis feeding into advice given to a client, risk-profiling assistance, document summarisation supporting an advisory recommendation, or a client-facing chatbot answering queries, Regulation 15(14)'s disclosure obligation applies at the agreement stage, and Regulation 16C's liability standard applies continuously, for as long as that tool remains in use. This is true regardless of whether the AI tool was built internally or licensed from a third party — the accountability sits with the IA either way.

What the Disclosure Itself Should Actually Cover

A defensible Regulation 15(14) disclosure goes beyond a single vague sentence acknowledging "we use technology in our process." It should specify, in plain language, which parts of the advisory process involve AI tools, the general purpose those tools serve, and — as SEBI's broader direction of travel (discussed below) increasingly signals — some indication of the tool's role relative to human judgment in the final advice given, so a client isn't left assuming a fully automated recommendation is actually a human-reviewed one, or vice versa.

Where SEBI's Thinking Is Headed — The June 2025 Consultation Paper

SEBI's June 2025 Consultation Paper on responsible AI/ML usage points toward a more comprehensive future framework for all regulated entities: board-approved AI governance policies, documented testing before deployment, ongoing monitoring, independent audit of AI systems, and client-facing disclosure covering purpose, risk, and accuracy in accessible language. None of this is binding law yet in the specific, detailed form the consultation paper describes — but IA practices building or expanding AI use now should treat this direction as a strong signal of where formal requirements are likely headed, rather than waiting for a further circular to force retrofitted compliance later.

Vendor Contracts Need a Second Look in Light of Regulation 16C

Because Regulation 16C places accountability squarely on the IA regardless of whose tool is being used, existing vendor agreements for any AI-powered tool in the advisory workflow should be reviewed specifically for clauses that attempt to disclaim the vendor's own liability for data handling or output quality — such clauses don't relieve the IA of its own regulatory accountability to SEBI, whatever they say about liability between the IA and the vendor commercially.

Building an Internal AI Usage Inventory

A practical first step for any IA practice using AI in any part of its operations is a simple, maintained inventory: what tools are in use, for what specific purpose, whether client data passes through them, and what human oversight sits around each tool's output before it reaches a client. This inventory is the foundation both for the Regulation 15(14) disclosure and for demonstrating Regulation 16C accountability if SEBI ever asks.

How We Help

We help Investment Advisers map their AI tool usage against current SEBI disclosure and liability requirements, draft the client-facing disclosure language Regulation 15(14) requires, review vendor agreements for the liability gaps Regulation 16C creates, and build the internal governance documentation SEBI's direction of travel is clearly pointing toward. If your practice uses AI anywhere in its advisory workflow and hasn't had this reviewed, our SEBI Intermediary Compliance practice can help you get ahead of it.

This article is for general informational purposes and reflects the regulatory position as understood at the time of writing. It does not constitute regulatory advice specific to any IA's circumstances — please consult our team and verify the current status of SEBI's AI/ML framework before relying on it for compliance decisions.