SEBI to Issue AI Guidelines for Markets
Why in News?
- On 10 October 2026, Securities and Exchange Board of India Chairman Tuhin Kanta Pandey said the regulator will shortly issue guidelines on the responsible use of AI and machine learning in capital markets.
- He spoke at the second Capital Market Confluence 2026, organised by the Bombay Stock Exchange Brokers' Forum.
- The framework will use a tiered approach with accountability requirements, data controls, kill switches and human oversight. It will align with the supervisory toolkit of the International Organization of Securities Commissions.
- "Technology cannot dilute accountability," he said. Regulated entities remain responsible for the outcomes of AI.
- SEBI had issued a consultation paper on AI and ML guidelines on 20 June 2025.
Key Terminologies
- AI and machine learning
- Software that learns patterns from data to make predictions or decisions. In markets it is used in areas such as algorithmic trading, asset management, portfolio management and advisory services.
- Kill switch
- A control that lets a human stop or switch off an AI system quickly when it behaves wrongly. SEBI named it as one of the safeguards in its planned framework.
- Tiered approach
- A rule design in which stricter requirements apply to higher-risk uses of AI. It avoids treating every tool the same way.
- IOSCO
- The International Organization of Securities Commissions, the global body of securities regulators. It published a Supervisory Toolkit for AI Use in Capital Markets on 25 May 2026.
Key Issues
- Accountability: SEBI says regulated entities stay responsible for AI outcomes. Its 20 June 2025 consultation paper proposed model governance, internal risk controls and oversight by senior management.
- Model Risk: The consultation paper asked that systems be tested and validated, and secure against malfunction and misuse. Kill switches in the new framework respond to the risk of an AI system acting wrongly at speed.
- Fairness: The consultation paper proposed principles of non-discriminatory outcomes and fair access to services. AI trained on skewed data could treat client groups differently.
- Transparency: The consultation paper proposed that intermediaries disclose AI and ML use in algorithmic trading, asset management, portfolio management, and advisory and research services, with records of development and monitoring.
- Regulatory Overlap: The Reserve Bank of India issued its FREE-AI report on 13 August 2025 with 26 recommendations for the financial sector. A firm regulated by both bodies may face two sets of expectations.
Key Implications
Positive/Pros/Merits
- Investor Protection: Human oversight and kill switches give investors a safeguard against automated errors, and the tiered approach focuses stricter rules on riskier uses.
- Global Alignment: IOSCO's toolkit has three layers, covering risk areas, oversight tools and monitoring indicators, so SEBI's alignment helps Indian firms operating across borders.
- Innovation Space: The IndiaAI Mission, approved by the Cabinet on 7 March 2024 with over Rs 10,372 crore over five years, builds AI capacity. Clear rules let firms use it with confidence.
Negative/Cons/Demerits
- Compliance Cost: Records of model development, testing and monitoring, and periodic reports, add costs. Smaller brokers and advisers may find this harder than large exchanges.
- Timing Uncertainty: SEBI says the guidelines will come shortly, but the content of the final tiers has not been published. Market participants cannot yet plan.
- Fast-Moving Technology: IOSCO notes that the AI lifecycle includes generative and agentic systems. A tiered rule written today may need frequent change.
Key Initiatives
- SEBI Consultation Paper, 20 June 2025: It proposed an AI and ML governance framework for securities markets covering model governance, investor protection, testing and disclosure.
- Reserve Bank of India FREE-AI Report, 13 August 2025: The committee set up on 26 December 2024 gave seven guiding principles and 26 recommendations for responsible AI in finance.
- IndiaAI Mission: The Union Cabinet approved it on 7 March 2024 with over Rs 10,372 crore over five years, with about 10,000 GPUs planned through a public-private partnership.
- IOSCO, 25 May 2026: The Supervisory Toolkit for AI Use in Capital Markets covers the full AI lifecycle and gives oversight tools on governance, third-party risk, disclosure and recordkeeping.
- European Union: The AI Act entered into force on 1 August 2024 and became broadly applicable on 2 August 2026, with risk-based categories and high-risk obligations from 2 December 2027.
SEBI says it will shortly issue AI and machine learning guidelines using a tiered approach, and that regulated entities remain accountable for AI outcomes. It is also reviewing comments on a consultation paper on expiry-day derivatives settlement, with guidance expected in about a week. The Reserve Bank of India has separately published its FREE-AI framework for the wider financial sector.
Director's Perspective
Way Forward
- Publish the final tiers with examples, so firms know which uses count as higher risk.
- Make kill switches testable, with logs that regulators can audit after an incident.
- Coordinate SEBI and RBI expectations through a common template for firms under both regulators.
- Give smaller intermediaries a phased timeline and standard record formats to limit costs.
SEBI's plan is sound in direction: accountability stays with the firm, and safeguards scale with risk. It matches the IOSCO toolkit and builds on its own 2025 consultation. The gaps are the missing final text and the overlap with the RBI's advisory framework. In a Mains answer, credit the risk-based, human-in-control design, then conclude that clear tiers and coordinated regulators will decide whether it protects investors without stifling innovation.
GS Relevance
Frequently Asked Questions
What did SEBI announce about AI on 10 October 2026?
SEBI Chairman Tuhin Kanta Pandey said on 10 October 2026 that the regulator will shortly issue guidelines on responsible use of AI and machine learning in capital markets. They will use a tiered approach with accountability, data controls, kill switches and human oversight.
What is a kill switch in AI regulation?
A kill switch is a control that lets humans stop an AI system quickly if it malfunctions. SEBI named kill switches, with human oversight and data controls, as part of its planned framework for AI and machine learning in Indian capital markets.
How is India regulating AI in finance?
India uses sector regulators: SEBI is preparing AI and ML guidelines for markets, and the Reserve Bank of India issued its FREE-AI report with 26 recommendations on 13 August 2025. The IndiaAI Mission, approved on 7 March 2024, supports AI capacity with over Rs 10,372 crore.
PYQ Practice — Statement Analysis
1 SEBI has said that regulated entities remain responsible for the outcomes of AI systems they use.
The Chairman said on 10 October 2026 that technology cannot dilute accountability.
2 The RBI FREE-AI report has seven guiding principles and 26 recommendations.
It was released on 13 August 2025 for the financial sector.
3 The IndiaAI Mission was approved by the Union Cabinet in 2019 with an outlay of Rs 1,000 crore.
It was approved on 7 March 2024 with over Rs 10,372 crore over five years.