Quick answer
SEBI regulates AI in the securities market through three layers: 2019 circulars requiring intermediaries, market infrastructure institutions and mutual funds to report AI/ML systems quarterly; an early-2025 amendment (regulation 16C, Intermediaries Regulations) making a regulated person solely responsible for AI outputs, data privacy and compliance regardless of vendor; and a June 2025 consultation on responsible AI/ML guidelines. Check whether final guidelines have been issued.
- SEBI's 2019 circulars require market intermediaries, market infrastructure institutions and mutual funds to report AI/ML systems to SEBI quarterly.
- Regulation 16C, added in early 2025, makes a regulated person using AI tools solely responsible for data privacy, security, outputs and compliance.
- Using a third-party AI vendor does not shift SEBI accountability away from the regulated entity.
- SEBI's June 2025 consultation paper proposed guidelines on governance, model testing, client disclosure, fairness and data security; check whether they are final.
- Investment advisers and research analysts using AI must disclose that use to clients.
Who this is for: Stock brokers and depository participants, Investment advisers and research analysts, Mutual funds and asset management companies, Stock exchanges, depositories and clearing corporations, Algo trading and wealth-tech platforms
How does SEBI regulate AI and ML?
SEBI does not have a single AI regulation. Its approach has built up in three layers.
| Layer | When | What it does |
|---|---|---|
| AI/ML reporting circulars | January and May 2019 | Market intermediaries, market infrastructure institutions (MIIs) and mutual funds report the AI/ML applications and systems they use, quarterly |
| Sole responsibility amendment | Early 2025 | Regulation 16C of the Intermediaries Regulations: a regulated person using AI tools is solely responsible for data privacy and security, outputs and compliance, whether the tool is built in-house or bought |
| Consultation on guidelines | June 2025 | Proposed guidelines for responsible AI/ML usage: governance, model testing, disclosure to clients, fairness and data security |
Separately, investment advisers and research analysts using AI tools must disclose that use to their clients.
The consultation paper is a proposal. Before relying on any specific requirement — for example, testing frequency or disclosure wording — check sebi.gov.in for whether final guidelines or a circular have been issued.
The rest of this page explains what each layer means for engineering and compliance teams, and how SEBI's approach meets the DPDP Act from 13 May 2027.
Quarterly AI/ML reporting under the 2019 circulars
The 2019 circulars were among the first AI-specific supervisory requirements in Indian finance. They ask regulated entities to tell SEBI which AI/ML applications they use, in a prescribed format, every quarter.
Who reports
- market intermediaries (such as stock brokers and depository participants);
- market infrastructure institutions (stock exchanges, depositories, clearing corporations);
- mutual funds and their asset management companies.
What a usable internal process looks like
- Maintain an AI/ML inventory that is the single source for SEBI reporting. Fields should include application name, business function, whether it is client-facing, the technique used, data sources, vendor and owner.
- Define what counts. The circulars define the applications in scope; read the definitions and decide consistently whether rules-based systems, vendor analytics and generative AI tools are included. When unsure, err on including.
- Assign a reporting owner in compliance, with technology sign-off.
- Reconcile quarterly against procurement, cloud billing and model registries to catch new tools.
The circulars predate generative AI. A broker using an LLM to summarise research or answer client queries should treat that as a reportable AI application unless SEBI guidance says otherwise.
Regulation 16C: you own the AI's output
The early-2025 amendment inserting regulation 16C into the SEBI (Intermediaries) Regulations is the most consequential layer. Its effect: a regulated person that uses AI tools is solely responsible for:
- the privacy, security and integrity of investor and stakeholder data used by the tool;
- the output of the tool and actions taken on it;
- compliance with applicable laws.
This applies irrespective of the scope of AI use or whether the tool is from a third party.
Practical consequences
- Vendor contracts cannot transfer regulatory risk. You can seek indemnities, but SEBI will hold the regulated entity accountable.
- Explainability matters. If you cannot explain why an AI tool generated a recommendation or flagged a trade, you will struggle to defend it.
- Human review is a control, not a formality. For client-facing outputs (research notes, advice, suitability), keep evidence of review.
- Data flows into vendor tools need governance. Sending client data to an AI API is a data security decision that SEBI may examine.
This is very close to the DPDP principle in section 8(1) that a Data Fiduciary remains responsible for processing done by its Data Processors.
The June 2025 consultation: what to prepare for
SEBI's June 2025 consultation paper proposed guidelines for responsible usage of AI/ML in the securities market. The themes it covered:
- Governance — board and senior management oversight, accountable officials, policies.
- Model testing — validation before deployment and monitoring after, including behaviour in stressed markets.
- Disclosure to clients — telling clients when AI materially affects services they receive.
- Fairness — avoiding outcomes that systematically disadvantage groups of investors.
- Data security — protecting data used to train and run models.
Final guidelines may differ from the consultation in scope, thresholds or timing, and may distinguish by entity size or use-case risk. Check whether final guidelines have been issued before building to specific requirements.
No-regret preparation
- Extend your AI/ML inventory with risk tiers.
- Document validation for each client-facing or trading model.
- Draft client disclosure language for AI-assisted advice and research.
- Set up fairness checks where models segment or price clients.
- Map data flows to vendors and apply DPDP-grade safeguards.
Banks and NBFCs face a parallel track under RBI's FREE-AI framework, which shares many of the same themes.
Investment advisers and research analysts
For investment advisers (IAs) and research analysts (RAs), the AI question is personal: advice and research carry the professional's name and registration.
Disclosure
IAs and RAs using AI tools in providing services must disclose that use to clients. Keep the disclosure specific: what the tool does (for example, drafts research summaries, screens portfolios), and that a registered person reviews and owns the output.
Accountability
Under regulation 16C-style responsibility, an AI-generated recommendation is the adviser's recommendation. Suitability, conflicts and record-keeping duties are unchanged.
Engineering patterns that help
- Draft-and-approve workflow. AI drafts; the registered person edits and approves; the system records the approver and timestamp.
- Source logging. Record the data sources used for each AI-assisted output.
- Prompt templates under change control. A change in template can change advice; treat it like a model change.
- Client data minimisation. Send the AI tool only what it needs. Use tokenised client IDs rather than names and PANs where possible.
For wealth-tech platforms that process client personal data at scale, DPDP for banks, NBFCs and fintechs covers the data protection side.
SEBI AI rules and the DPDP Act
SEBI's rules tell you who is accountable for AI in market activities. DPDP tells you how to treat the personal data inside those systems from 13 May 2027.
| SEBI expectation | DPDP obligation |
|---|---|
| Data privacy and security of AI tools (reg. 16C) | Rule 6 reasonable security safeguards: encryption/masking, access control, logs, backups, processor contract clauses |
| Responsibility for vendor tools | s.8(1): fiduciary responsible for processing by processors; s.8(2) valid contract |
| Client disclosure of AI use | Rule 3 notice of personal data and purposes |
| AI/ML inventory | Data mapping for rights requests under Rule 14 |
| Incident handling | Rule 7 breach intimation; detailed Board report within 72 hours |
| Record-keeping | Rule 8(3) one-year minimum retention of personal data, traffic data and logs |
Where SEBI or other securities law requires longer retention, that law governs, and DPDP's erasure duty under section 8(7) yields to retention required by law.
Common mistakes
- Reporting only "machine learning" systems and omitting LLM tools.
- Assuming a vendor's certification satisfies regulation 16C.
- Sending full client KYC data to an external model for convenience.
- Treating the June 2025 consultation as final.
Further reading on suhasbhairav.com
- BlogPrinciples-Based Governance vs Operational Controls for Responsible AI
- BlogBias Evaluation vs Fairness Auditing in Production AI
- BlogAuditing AI Decisions: End-to-End Governance in Production
- AI AutomationAI Automation for Compliance Surveillance in Capital Markets
- ResearchLLM Leaderboard & AI Model Benchmarks
Before 13 May 2027
Compliance checklist
Maintain an AI/ML inventory
Keep one register of AI/ML applications that feeds SEBI quarterly reports.
SEBI 2019 AI/ML circulars
File quarterly AI/ML reports
Assign an owner and reconcile against procurement and model registries each quarter.
SEBI 2019 AI/ML circulars
Accept sole responsibility in governance
Document that AI outputs are owned by named officials, irrespective of vendor.
Reg. 16C, Intermediaries Regulations
Review AI vendor data flows
Minimise client data sent to AI tools and add security and audit clauses.
Reg. 16C; DPDP Rule 6(1)(f)
Disclose AI use to clients
Investment advisers and research analysts must tell clients how AI is used.
SEBI IA/RA requirements
Validate client-facing models
Test before deployment and monitor afterwards; keep evidence.
SEBI June 2025 consultation (proposed)
Check for final SEBI guidelines
Monitor sebi.gov.in for final responsible AI/ML guidelines and update controls.
SEBI June 2025 consultation
Align incident handling with DPDP
Route AI incidents involving personal data into the Rule 7 workflow.
DPDP Rule 7
Key dates
DPDP core obligations apply to personal data in SEBI-regulated AI systems
Answers
Frequently asked questions
Does SEBI regulate AI in stock broking?
Does SEBI regulate AI in stock broking?
Yes. SEBI's 2019 circulars require market intermediaries, market infrastructure institutions and mutual funds to report AI/ML systems quarterly. An early-2025 amendment, regulation 16C of the Intermediaries Regulations, makes regulated persons solely responsible for AI outputs, data privacy and compliance, and a June 2025 consultation proposed fuller guidelines.
What is SEBI regulation 16C?
What is SEBI regulation 16C?
Regulation 16C of the SEBI (Intermediaries) Regulations, added in early 2025, says a regulated person using AI tools is solely responsible for the privacy and security of data, the output of the tools, and compliance with applicable law, regardless of whether the tool is in-house or from a third-party vendor.
Has SEBI issued final AI guidelines?
Has SEBI issued final AI guidelines?
SEBI issued a consultation paper in June 2025 on guidelines for responsible AI/ML usage in the securities market. Check sebi.gov.in for whether final guidelines or a circular have been issued, because the final version may change scope, thresholds and timelines from the consultation.
Do investment advisers have to disclose AI use?
Do investment advisers have to disclose AI use?
Yes. Investment advisers and research analysts who use AI tools in their services must disclose that use to clients. The adviser remains responsible for the advice, so disclosure should explain what the tool does and that a registered person reviews and owns the output.
Can a broker blame its AI vendor for a wrong output?
Can a broker blame its AI vendor for a wrong output?
Not with SEBI. Regulation 16C places sole responsibility on the regulated person regardless of the vendor. You can seek contractual remedies from the vendor, but regulatory accountability stays with you. DPDP section 8(1) takes a similar position for personal data processed by processors.
Does DPDP apply to SEBI-regulated entities?
Does DPDP apply to SEBI-regulated entities?
Yes. Brokers, advisers, mutual funds and MIIs processing digital personal data of individuals are Data Fiduciaries. From 13 May 2027 they must meet DPDP Rules on notice, security, breach intimation, retention and rights, alongside SEBI requirements. Where securities law requires longer retention, it prevails.
Legal references and sources
- SEBI circulars on reporting of AI/ML applications — January and May 2019 circulars (official text)
- SEBI (Intermediaries) Regulations, 2008 — Regulation 16C (official text)
- Digital Personal Data Protection Rules, 2025 — Rules 6 and 7 (official text)
This guide explains the law for engineering and business teams. It is not legal advice; confirm how it applies to your organisation with qualified counsel, and check the Gazette for later notifications or amendments.
About the author

Suhas Bhairav
Systems architect and applied AI practitioner who designs production AI systems for Indian and global companies, with a focus on privacy-by-design under the DPDP Act and the EU AI Act.
Related guides
- India AI & sector rulesRBI FREE-AI Framework: What Banks & Fintechs Must DoWhat the RBI FREE-AI committee report asks of banks, NBFCs and fintechs, what is and is not binding yet, and how it maps onto DPDP obligations from May 2027.
- Sector guidesDPDP for Banks, NBFCs and FintechsHow banks, NBFCs, payment companies and fintechs layer DPDP obligations on top of RBI rules: consent, retention, breach, cross-border, AI credit models and Consent Managers.
- India AI & sector rulesIndia AI Regulation in 2026: The Complete LandscapeIndia regulates AI through existing laws, not a standalone AI Act: MeitY's AI Governance Guidelines, DPDP, IT Rules on synthetic content, CERT-In directions, RBI FREE-AI and SEBI rules.
- AI & personal dataAI Vendor Due Diligence Checklist Under the DPDP ActA practical due diligence checklist for LLM APIs, vector databases, observability tools and AI SaaS vendors, mapped to DPDP section 8 and Rules 6, 7 and 8(3).
- India AI & sector rulesCERT-In Rules for AI Systems: Incident Reporting & LogsHow CERT-In's April 2022 directions on 6-hour incident reporting, 180-day logs in India and clock sync apply to LLM gateways, model APIs and AI agents, and how they meet DPDP.
- India AI & sector rulesAI Labelling in India vs EU AI Act Article 50Side-by-side comparison of India's IT Rules synthetic content labelling and EU AI Act Article 50 transparency duties, with a single build plan that can satisfy both.