AI in capital markets is the use of machine learning, natural language processing (NLP) and generative models in research, trading, operations and supervision. In India it now runs earnings call transcription, monitoring, documentation checks and client reporting at advisers, brokers and asset management companies (AMCs), while the Securities and Exchange Board of India (SEBI) holds firms responsible for every output.
Key points
- In a Nextyn Consulting & Research interview in December 2025, a former partner at an Indian investment advisory firm estimated that AI tools can cut the people needed for routine research and operations work by 20% to 25%.
- The same interview found that one analyst using live transcription and summaries can cover roughly twice as many companies in an earnings season.
- The former partner would not hand investment decisions to current AI: "I would not trust my funds with an AI."
- SEBI chairman Tuhin Kanta Pandey said on 19 August 2026, at the Federation of Indian Chambers of Commerce and Industry (FICCI) Capital Markets Conference in Mumbai, that forthcoming AI guidelines would require kill switches, humans in the loop and data controls.
- The National Stock Exchange of India (NSE) counted about 13.5 crore (135 million) unique registered investors on 31 August 2026, and Central Depository Services (India) Limited (CDSL) and National Securities Depository Limited (NSDL) held about 24.4 crore (244 million) demat accounts at the end of October 2026.
Method: this article draws on a Nextyn Consulting & Research interview, 17 December 2025, with a former partner at an Indian investment advisory firm serving family offices and ultra high net worth individuals (UHNIs), plus public regulatory and market data as of 16 November 2026.
How is AI used in capital markets in India today?
In India, AI works mostly as a research companion and an operations tool. Firms apply it to live transcription of earnings calls, summaries of filings, portfolio reporting, compliance checks and monitoring alerts. The former advisory partner interviewed by Nextyn Consulting & Research sees little evidence yet that AI produces investment decisions worth acting on without human analysis.
Discretionary and quant managers alike still start from a human idea, and AI fine tunes the work around it. Across financial services, its clearest value lies in tasks such as flagging pending audit items.
Where AI equity research adds the most
AI equity research earns its keep through coverage. Live transcription and NLP summaries let one analyst follow several overlapping earnings calls at once. The former advisory partner estimates this roughly doubles the number of companies an analyst can track, though with less depth outside the core watch list.
An analyst covering 100 earnings calls in a season spends about 100 hours listening. The alternative described in the interview: attend the key call live while three others are transcribed and summarized in parallel.
AI investment research changes desk economics. The former partner estimates a task needing four people can be handled by three, because one person with good prompts processes far more material. The tools named were Perplexity, Claude, Gemini and NotebookLM.
Why AI financial statement analysis still needs an analyst
AI financial statement analysis remains the weakest link. In the former partner's experience, current tools largely restate what is already published and do not reliably read a balance sheet or profit and loss statement for the details that drive value, such as cross holdings, contingent liabilities or unusual line items.
The interview offered an example: Tata Chemicals holds a stake in Tata Sons that several analysts have valued close to, or above, Tata Chemicals' own market capitalization, depending on the Tata Sons valuation assumed. A general model asked about the stock may miss that holding.
General AI tools show the same limit on SEBI regulation: a tool may cite the latest circular and miss the master circular it amends, and the former partner has seen wrong compliance answers as a result. AI for financial analysis is a useful first pass; the compliance call rests with the person accountable for the data.
Operational intelligence: automation first, then real time monitoring
Operational intelligence is the continuous analysis of operational data so a firm can act on events as they happen. In Indian capital markets it starts with plain automation: intermediaries first automate core processing, such as contract notes that arrive from exchanges between roughly 8 pm and 10 pm, and only then add AI for monitoring and alerts.
Larger institutions, already automated, are moving to the next layer, mirroring the shift from AI pilots to production in Indian banking. Smaller intermediaries are still automating core processing and often prefer rule based processes, a path Indian banks took in core banking modernization.
The next layer is real time operational intelligence: alerts when a defined dataset changes, risk management dashboards across asset classes, and explainable alerts. BlackRock's Aladdin, which combines risk, allocation and cross holding views, was described in the interview as the benchmark Indian builders aim for.
SEBI is applying AI to its own supervision: its Annual Report 2025 to 26 describes Project Sudarsan, an AI platform that tracks unsolicited advice on social media, and R(AI)DAR, which reviews AMC advertisements. Similar tools appear in the fraud and risk technology landscape.
| Function | Current AI role | Maturity in India (interview view) | Where the human stays |
|---|---|---|---|
| Earnings call coverage | Live transcription, NLP summaries | Widely used | Choosing what matters |
| Financial statement review | First pass extraction | Limited reliability | Valuation and accounting judgment |
| Compliance and audit | Pending item checks, circular search | Useful with caution | Reading master circulars and addenda |
| Portfolio reporting | Queries across family holdings | Practical today | Client conversation and suitability |
| Monitoring and risk | Alerts on data changes, dashboards | Growing in large firms | Setting thresholds, acting on alerts |
| Execution | Bots that suggest and place orders | Live at some brokers, modest value | Trade intent and limits |
AI in trading: broker bots, retail algorithms and surveillance
AI in Indian trading shows up in three places: broker bots that suggest and place orders, rule based algorithms that AI helps code and test, and market surveillance. The former advisory partner sees execution bots, such as one that proposes a Reliance trade and offers to place it, as a small convenience rather than better advice.
Machine learning for asset management is most visible in quant funds, whose rules still come from a person. AI helps code those rules, back test them and link them to a broker. "AI is going to prepare the language for you, but the input still remains yours," the former partner said.
SEBI's 4 February 2025 circular on safer retail participation in algorithmic trading became fully mandatory for all stock brokers on 1 April 2026. On market surveillance, NSE's Chief Business Development Officer said in September 2026 that the exchange uses homegrown AI algorithms to detect manipulation and insider trading.
Build or buy: AI in asset management
AI in asset management is splitting by firm size. Large regulated managers and custodians build in house models on their own servers to keep control of client data. Smaller advisers and retail brokers use third party AI tools on an ad hoc basis, a choice that technology scouting can test against building.
The build or buy split is driven by regulation. Public money brings public data, as the former partner put it, and no firm wants a regulatory dispute over where it is stored. For a depository such as CDSL, cost is not the constraint; the question is whether a clear use case exists.
Among retail brokers, Zerodha launched Kite MCP (Model Context Protocol) in May 2025, letting clients connect accounts to assistants such as Claude. Groww's 23 October 2026 product update says its GR1 assistant shows insights for 750 stocks and never places an order without explicit confirmation. The interview framed these as retention tools: if even 10% of active clients want them, a broker has to offer them.
| Segment | Typical approach | Main AI uses | Primary constraint |
|---|---|---|---|
| Large AMCs, custodians, depositories | Build in house, data on own servers | Monitoring, risk alerts, reporting | Data residency, SEBI responsibility |
| Hedge funds and quant managers | In house models, AI assisted coding | Rule coding, back testing, execution links | Quality of human input, model risk |
| Advisers to UHNIs and family offices | Third party tools, ad hoc | Research summaries, reporting, audit checks | Accuracy on compliance and valuation |
| Retail brokers | Embedded assistants, agentic features | Queries, order suggestions, account insights | Investor protection, suitability, retention |
What SEBI says about AI use by market intermediaries
SEBI holds every regulated entity fully responsible for the AI tools it uses, whether built in house or bought. Research analysts must disclose AI use to clients, and 2025 amendments extend responsibility for data and outputs to all intermediaries. As of 16 November 2026, final guidelines with kill switches were announced but not issued.
December 2024 amendments to the SEBI Research Analysts Regulations required disclosure of the extent of AI use and made analysts responsible for advice based on AI output. In February 2025, SEBI amended the Intermediaries Regulations, so that any regulated entity is solely responsible for investor data and AI output.
SEBI's consultation paper of 20 June 2025 proposed principles on model governance, disclosure, testing, fairness and data security. On 19 August 2026, in Mumbai, the chairman said the final AI governance rules would add kill switches, humans in the loop and data controls.
SEBI's approach to AI in capital markets tracks global supervisors. The International Organization of Securities Commissions (IOSCO) published its consultation report "Artificial Intelligence in Capital Markets: Use Cases, Risks, and Challenges" in March 2025. The International Monetary Fund (IMF) October 2024 Global Financial Stability Report found AI could make markets more efficient but also more opaque and volatile.
The former partner reads SEBI's intent through its investor protection mandate and expects little appetite for faceless advisory. SEBI's August 2026 study found that 87.7% of individual futures and options (F&O) traders lost money in FY26 (April 2025 to March 2026), against 90.9% in FY25, with net losses of about ₹91,685 crore (about ₹917 billion).
Banks face a parallel regime: the Reserve Bank of India (RBI) published its Framework for Responsible and Ethical Enablement of Artificial Intelligence (FREE AI) report on 13 August 2025.
A practical sequence for AI portfolio management
The sequence recommended in the interview starts with automation, not models. Complete automation of core processing, then fund secure in house pilots for monitoring, explainable alerts and tightly limited agentic execution. AI portfolio management should support the manager's judgment, and data governance deserves a material share of spend.
For firms building AI in investment management, each use case should earn funding on its own evidence. Decision support will follow as accuracy improves, so AI for capital markets will deepen first where a human is accountable for every output.
- 01
Automate core processing
Clear contract notes and back office work before adding AI.
- 02
Pilot monitoring in house
Keep client data on own servers for alerts and dashboards.
- 03
Document every model
Record testing, owners and kill switch controls for SEBI.
- 04
Measure each use case
Fund only uses with evidence, starting with transcription and reporting.
What this means for AMCs, brokers, advisers and investors
AI in capital markets in India rewards firms that automate first, keep client data close and document every model for SEBI. The investment decision itself stays human, so priorities differ by segment for the asset managers, brokers, advisers and investors most affected by these shifts.
- AMCs and custodians: Build AI investment management capability close to your own data, starting with monitoring and explainable alerts, and document each model for SEBI's governance and kill switch expectations.
- Brokers on NSE and BSE Limited: Automate contract notes and back office work before adding AI, and treat retail AI features as retention tools with suitability controls.
- Advisers and research analysts: Use AI to widen coverage and speed reporting, disclose that use as SEBI requires, and keep valuation and compliance calls in human hands.
- Investors and allocators: Ask managers which tasks AI performs, how outputs are checked, and where client data, including any alternative data, is stored.
AI in capital markets in India is already changing how research is produced and how operations run, while the investment decision remains where experience counts most. Nextyn Consulting & Research tests AI build, buy and investment theses with primary research and regulation impact analysis, including interviews with practitioners who have built these workflows in Indian advisory, broking and asset management firms.
Sources
- Securities and Exchange Board of India (SEBI), "Annual Report 2025-26", Chairman's Statement and Chapter 9 (Project Sudarsan, R(AI)DAR), August 2026. https://www.sebi.gov.in/reports-and-statistics/publications/aug-2026/Chairmans%20Statement.pdf and https://www.sebi.gov.in/reports-and-statistics/publications/aug-2026/Chapter%2009.pdf
- SEBI, "Profitability of Individual Traders in the Equity Derivatives Segment (FY25 to FY26)", August 2026. https://www.sebi.gov.in/sebi_data/attachdocs/aug-2026/1787233506209.pdf
- SEBI, circular extending the timeline for "Safer participation of retail investors in Algorithmic trading" (original circular 4 February 2025), 30 September 2025. https://www.sebi.gov.in/sebi_data/attachdocs/sep-2025/1759232056254.pdf
- SEBI, board memorandum "Assigning responsibility for the use of artificial intelligence tools by market infrastructure institutions, registered intermediaries and other persons regulated by SEBI", December 2024. https://www.sebi.gov.in/sebi_data/meetingfiles/dec-2024/1735042007618_1.pdf
- AZB and Partners, "SEBI introduces provisions in relation to usage of artificial intelligence and machine learning tools by market infrastructure institutions and registered intermediaries", 2025. https://www.azbpartners.com/bank/sebi-introduces-provisions-in-relation-to-usage-of-artificial-intelligence-and-machine-learning-tools-by-market-infrastructure-institutions-and-registered-intermediaries/
- SCC Online, "SEBI (Research Analysts) (Third Amendment) Regulations, 2024", December 2024. https://www.scconline.com/blog/post/2024/12/19/sebi-research-analyst-third-amendment-regulations-2024/
- Taxmann, "SEBI proposes AI/ML governance framework for securities markets" (SEBI consultation paper of 20 June 2025), 2025. https://www.taxmann.com/post/blog/sebi-proposes-ai-ml-governance-framework-for-securities-markets
- Outlook Money, "SEBI to issue AI/ML guidelines for capital markets: mandatory kill switch and human oversight framework explained" (FICCI Capital Markets Conference, Mumbai, 19 August 2026), 2026. https://www.outlookmoney.com/invest/sebi-to-issue-aiml-guidelines-for-capital-markets-mandatory-kill-switch-and-human-oversight-framework-explained
- International Organization of Securities Commissions (IOSCO), "Artificial Intelligence in Capital Markets: Use Cases, Risks, and Challenges", consultation report CR/01/2025, March 2025. https://www.iosco.org/library/pubdocs/pdf/IOSCOPD788.pdf
- International Monetary Fund (IMF), "Global Financial Stability Report, October 2024", Chapter 3 "Advances in Artificial Intelligence: Implications for Capital Market Activities", 2024. https://www.imf.org/en/Publications/GFSR/Issues/2024/10/22/global-financial-stability-report-october-2024
- Reserve Bank of India (RBI), press release on the FREE AI Committee report, 13 August 2025. https://www.rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=61018
- National Stock Exchange of India (NSE), "Market Pulse", September 2026 (13.5 crore unique registered investors at 31 August 2026). https://nsearchives.nseindia.com//web/mediaattachment/2026-09/Market_Pulse_September_2026_Final.pdf
- Central Depository Services (India) Limited (CDSL), "Periodic Details for Month of October 2026" (about 19.63 crore BO accounts at month end), November 2026. https://www.cdslindia.com/Downloads/Publications/Periodic%20Stats/October-2026.pdf
- National Securities Depository Limited (NSDL), "NSDL at a Glance" statistics (about 4.77 crore client accounts at 31 October 2026), 2026. https://nsdl.com/about-us/statistics
- WION, "National Stock Exchange of India using AI to track and curb malpractices", 12 September 2026. https://www.wionews.com/india-news/national-stock-exchange-of-india-using-ai-to-track-curb-malpractices-1789228630943
- Zerodha, "Connect your Zerodha account to AI assistants" (Kite MCP), 2025. https://zerodha.com/products/mcp/
- Groww, "GR-1 AI investing assistant" product update, 23 October 2026. https://groww.in/updates/gr-1-ai-investing-assistant-groww
- Outlook Business, "Tata Sons IPO: the listing that could reprice the Tata empire", 2026. https://www.outlookbusiness.com/markets/tata-sons-ipo-the-listing-that-could-reprice-the-tata-empire