What AI governance obligations does the SEBI Master Circular for Portfolio Managers 2025 create for PMS providers?

SEBI Master Circular for Portfolio Managers (SEBI/HO/IMD/IMD-POD-1/P/CIR/2025/104, issued July 16, 2025) supersedes the previous master circular dated June 7, 2024 and consolidates all directives up to March 31, 2025. Crucially, it incorporates the February 2025 SEBI (Intermediaries) (Amendment) Regulations Chapter III-A AI obligations into the portfolio management sector's comprehensive compliance framework. Five AI governance obligations for portfolio managers: (1) Mandatory AI/ML disclosure in the Disclosure Document filed with SEBI: portfolio managers must disclose AI/ML tool usage in the Disclosure Document — the principal client-facing and SEBI-filed document that defines the PMS strategy, risks, fees, and investment process. The disclosure must cover: (a) AI tools used in portfolio construction, stock selection, factor analysis, or execution; (b) risks and limitations of AI/ML systems including model risk and data dependency; (c) accuracy assumptions and data quality considerations; (d) the role of AI in discretionary versus automated decision-making within the PMS strategy; (2) Sole non-delegable liability for AI-driven investment decisions: portfolio managers bear sole responsibility for all investment decisions made or influenced by AI/ML tools — including quantitative factor models, momentum algorithms, and AI-driven portfolio rebalancing. Liability does not transfer to AI model vendors, data providers, or technology partners; (3) SEBI reporting for all AI/ML systems deployed: all AI/ML systems used in portfolio management must be reported to SEBI through the supervisory tracking mechanism. This includes in-house quantitative models, licensed third-party factor models, and any AI trading infrastructure; (4) Full compliance with DPDP Act 2023 for AI-processed client data: HNI client portfolio data, investment objectives, financial positions, and mandate terms — processed through AI for portfolio construction or rebalancing — must comply with DPDP Act purpose limitation, data minimisation, and retention requirements; (5) SEBI cybersecurity framework compliance: AI systems deployed in portfolio management must comply with SEBI's cybersecurity and cyber resilience framework applicable to intermediaries — covering AI model security, access controls, and incident reporting.

Why does AI governance matter specifically for India's SEBI-registered Portfolio Managers and quantitative PMS strategies?

India's PMS sector has grown to approximately 400+ registered portfolio managers managing over ₹30 lakh crore (approximately USD 360 billion) in assets for high-net-worth individuals, family offices, and qualified investors. AI and quantitative strategies have been one of the fastest-growing segments within PMS. Five reasons why AI governance is particularly significant for the PMS sector: (1) Discretionary versus systematic AI: PMS strategies range from fully discretionary (fund manager makes all decisions) to systematic/quantitative (AI model drives all portfolio decisions) to hybrid (AI provides signals, fund manager has final discretion). SEBI's AI disclosure obligation requires PMS providers to clearly characterize which category they fall into — and the liability framework is the same regardless of how much of the decision is AI-driven versus human; (2) Client mandate concentration: PMS clients have individual mandates with specific risk profiles, return objectives, and investment constraints. AI models that generate portfolio construction signals must be validated against the specific constraints of each client's mandate — a model that maximizes factor returns without respecting client-specific liquidity constraints or concentration limits is generating non-compliant recommendations regardless of AI accuracy; (3) Model risk in quantitative strategies: factor models, momentum strategies, and mean-reversion algorithms deployed in quantitative PMS carry model risk that differs from traditional discretionary fund management. In a market drawdown that reverses established factor relationships, a purely AI-driven PMS strategy may generate significant client losses before human intervention can override the model. SEBI's AI disclosure requirement forces PMS providers to document these risks in the Disclosure Document; (4) Leverage and derivatives: many quantitative PMS strategies use derivatives or leverage to amplify factor exposures. AI models governing leveraged positions carry higher-order risk than equity-only strategies — the accountability framework for AI governance of leveraged PMS strategies requires robust human oversight documentation; (5) Benchmark and performance attribution: clients of quantitative PMS strategies need to understand whether outperformance is attributable to AI model alpha or to factor exposure. SEBI's disclosure requirements for AI tools must include enough transparency for a sophisticated HNI client to understand the source of returns and risks — preventing AI opacity from obscuring risk-taking.

How does the SEBI Portfolio Managers AI obligation interact with the SEBI Intermediaries Amendment Regulations 2025 Chapter III-A?

The SEBI Master Circular for Portfolio Managers 2025 explicitly consolidates the SEBI (Intermediaries) (Amendment) Regulations 2025 Chapter III-A obligations into the portfolio management compliance framework. This creates a layered AI governance requirement for portfolio managers: (1) Chapter III-A (binding from February 10, 2025): establishes the foundational AI accountability structure — sole non-delegable liability for AI outputs, data privacy/security/integrity, full applicable law compliance, scale-independence. These obligations apply to portfolio managers as SEBI-registered intermediaries from February 10, 2025; (2) Master Circular consolidation (effective July 16, 2025): the Master Circular incorporates Chapter III-A into the portfolio management sector's comprehensive compliance framework, adding sector-specific implementation requirements — the Disclosure Document AI disclosure, SEBI reporting obligations, and cybersecurity framework compliance specific to portfolio management; (3) Disclosure Document specificity: the Disclosure Document requirement is PMS-specific — neither Chapter III-A nor the SEBI AI/ML Responsible Usage Guidelines establish a Disclosure Document AI disclosure requirement. This is a portfolio manager-specific obligation that applies on top of the general intermediary obligations; (4) Interaction with SEBI AI/ML Responsible Usage Framework (August 2026 final circular): portfolio managers must also comply with the August 2026 final circular when it became effective — adding kill-switch, human oversight, and vendor contract clause requirements on top of Chapter III-A's liability framework and the Master Circular's disclosure requirements. The compliance stack for a quantitative PMS provider is Chapter III-A + Master Circular AI obligations + Responsible Usage Framework operational requirements — all three applying simultaneously.

What must a SEBI-registered Portfolio Manager disclose about AI in its Disclosure Document to comply with the 2025 Master Circular?

SEBI's PMS Disclosure Document is the primary client-facing regulatory document for portfolio managers — it is filed with SEBI, provided to every client before engagement, and must be kept current. Under the 2025 Master Circular's AI governance obligations, the Disclosure Document must include comprehensive AI tool disclosure covering six dimensions: (1) AI strategy characterization: clear description of whether the strategy is fully discretionary (no AI in decision-making), AI-assisted (AI signals inform human decisions), AI-supervised (AI makes decisions subject to human override), or fully systematic (AI makes all allocation decisions without human intervention per trade); (2) AI tools and models used: identification of AI tools used in the investment process — proprietary quantitative models, licensed factor models, AI-driven execution algorithms, AI in portfolio monitoring and risk management. For third-party AI models, the Disclosure Document must disclose that they are third-party tools while noting the PM's sole liability; (3) AI-specific risk factors: a dedicated AI risk section covering model risk (risk that the AI model underperforms its historical validation results in live trading), data risk (risk that training data does not represent future market conditions), overfitting risk (risk that the model has been over-tuned to historical data), and concentration risk (risk that AI models generate correlated positions that increase portfolio drawdown in adverse conditions); (4) AI accuracy and data quality limitations: honest characterization of the AI model's known limitations — accuracy bounds, data sources and their vintage, out-of-sample performance versus backtested performance, performance under stress conditions; (5) Human oversight structure: description of the human oversight mechanisms over AI decisions — who reviews AI recommendations, at what frequency, under what criteria, and what triggers human override of AI-generated recommendations; (6) AI tool update and retraining policy: disclosure of how frequently AI models are retrained or updated, and what governance process governs model changes — so clients understand when the investment approach has materially changed due to AI model updates.

AI Governance · SEBI · Portfolio Managers · PMS · AI Disclosure · Quantitative Strategies · India BFSI

SEBI Portfolio Managers Master Circular 2025 AI Governance: Mandatory AI/ML Disclosure in PMS Disclosure Document, Sole Non-Delegable Liability for AI-Driven Portfolio Decisions and SEBI Reporting for 400+ PMS Providers Managing ₹30 Lakh Crore India

SEBI Master Circular for Portfolio Managers (SEBI/HO/IMD/IMD-POD-1/P/CIR/2025/104, July 16, 2025) supersedes the June 2024 master circular and consolidates the February 2025 SEBI Intermediaries Amendment Regulations Chapter III-A AI obligations into the PMS compliance framework. All 400+ SEBI-registered Portfolio Managers managing ₹30 lakh crore for HNIs, family offices, and qualified investors must: disclose AI/ML tool usage in the SEBI-filed Disclosure Document (covering tools, risks, limitations, and role in discretionary vs. systematic decision-making); bear sole non-delegable liability for AI-driven investment decisions; report all AI/ML systems deployed to SEBI; comply with DPDP Act for AI-processed client portfolio data. India's rapidly growing quantitative PMS sector — factor models, momentum algorithms, AI-driven rebalancing — faces a three-layer AI governance compliance stack: Chapter III-A + PMS Master Circular obligations + SEBI Responsible Usage Framework. Zero global competitor coverage.

Zytra Research Team · July 2025 · 11 min read

📖 Part of the India BFSI AI Governance Complete Guide 2026 — RBI, SEBI, IRDAI, NPCI & DPDP in one place.

A quantitative PMS provider that deploys AI for stock selection and portfolio rebalancing has a different AI governance challenge than a discretionary fund manager using AI only for data analysis. SEBI's disclosure obligation recognizes this — the Disclosure Document AI section must characterize whether the strategy is AI-assisted, AI-supervised, or fully systematic. That characterization determines the risk disclosure requirements and the oversight documentation required to satisfy the Master Circular.

The three-layer compliance stack is the practical challenge for quant PMS providers. Chapter III-A established the liability structure in February 2025. The Master Circular consolidated sector-specific disclosure and reporting in July 2025. The Responsible Usage Framework added operational requirements in August 2026. Each layer is cumulative — a PMS provider that complies with the most recent layer without addressing the earlier ones is still non-compliant with the earlier instruments.

The Disclosure Document Update Obligation

SEBI requires PMS providers to update the Disclosure Document whenever there are material changes to the strategy or investment process. An AI model retrain that materially changes the factor weights or strategy behaviour is likely a material change requiring a Disclosure Document update and potentially a fresh client communication. Quant PMS providers need to define what constitutes a material AI model change that triggers a Disclosure Document update — and maintain that definition consistently with SEBI's expectations.

SEBI Portfolio Managers AI Governance Compliance Checklist

☐ Disclosure Document AI section: current Disclosure Document includes AI/ML disclosure covering strategy characterization, tools used, AI-specific risk factors, accuracy limitations, human oversight structure, and model update policy
☐ Chapter III-A compliance: AI tool inventory completed; sole liability for AI outputs documented; vendor contracts reviewed; applicable law compliance mapping done — as required by SEBI Intermediaries Amendment Regulations 2025
☐ SEBI AI reporting: all AI/ML systems deployed in portfolio management reported to SEBI per supervisory tracking framework; reporting current and not deferred
☐ DPDP Act compliance for client portfolio AI: AI models processing HNI client data subject to DPDP Act purpose limitation, data minimisation, and retention policies; consent obtained for AI-assisted portfolio management
☐ Responsible Usage Framework: SEBI AI/ML Responsible Usage Framework operational requirements (kill-switch, human oversight documentation, vendor AI contract clauses) implemented for portfolio management AI systems
☐ Material change policy: definition of material AI model change that triggers Disclosure Document update documented; process for client communication on material AI changes defined
☐ SEBI inspection readiness: Disclosure Document with current AI section, SEBI AI reporting records, Chapter III-A compliance documentation — packaged for SEBI PMS inspection