SEBI calls for responsible, ethical use of AI in securities market; floats consultation paper

Shares acquired in FY25 but credited in FY26, which year the acquisition will be counted? SEBI clears the air in informal guidance.

The Securities and Exchange Board of India (SEBI) has floated a consultation paper seeking stakeholder views to ensure the responsible use of Artificial Intelligence (AI) and Machine Learning (ML) in the securities market.

In the consultation paper, SEBI has proposed guiding principles for the responsible use of AI and ML. “These guiding principles are intended to optimise benefits and minimise potential risks associated with the integration of AI/ML based applications in securities markets to safeguard investor protection, market integrity, and financial stability,” the paper said.

Also read: Shares acquired in FY25 but credited in FY26, which year the acquisition will be counted? SEBI clears the air in informal guidance. 

The guiding principles proposed in the consultation paper are Model Governance, Investor Protection-Disclosure, Testing framework, Fairness and bias, Data Privacy and Cyber Security measures.

Model governance

Under the proposed model governance, market participants using AI/ML models should have a skilled team to monitor the performance and oversee controls, testing, efficacy, and security of the algorithms. They are required to have appropriate risk control measures and governance frameworks, especially during market stress.

They need to appoint designated senior management having appropriate technical knowledge and experience, who would be responsible for the oversight of the development, validation, ongoing testing, deployment, monitoring and controls of AI/ML based models. Monitoring of third-party vendors and sharing of the accuracy of AI/ML models on a periodic basis to SEBI has also been stipulated.

Investor protection and disclosures

Market participants using AI/ML models for business operations that may impact their customers have to disclose the same to their customers to foster trust, transparency and accountability, according to the consultation paper.

The participants would need to disclose parameters as to how the trading algorithms have been selected, how it is being used for asset management/portfolio management, usage in advisory and support services, etc. The market regulator has given a non-exhaustive list where disclosure is required if AI/ML is used.

Testing framework

The SEBI paper suggests that market participants should adequately test and monitor the AI/ML based models to validate results on a regular basis. The testing should be done in a segregated environment from the live environment prior to deployment in order to ensure that AI/ML models behave as expected in stressed and unstressed market conditions.

The participants need to maintain proper documentation of all the models and store input and output data for at least 5 years. Proper documentation explaining the logic of AI/ML models is also required to ensure that the outcomes produced are explainable, traceable and repeatable. Continuous testing of AI/ML models has also been suggested.

Fairness and bias

The consultation paper said that AI/ML based models should be fair. Specifically, they should not favour or discriminate against one group of customers over another. Market participants should ensure that the data used is of adequate quality and sufficiently broad. This can be done by checking the quality of the sources used, as well as the relevance and completeness of the data about the objectives of the model. There should be appropriate processes and controls to identify and remove biases from data sets, the paper said.

Data privacy and cybersecurity

The SEBI paper has proposed that as AI/ML systems are dependent on collection and processing of data, market participants should have a clear policy for data security, cyber security and data privacy for the usage of AI/ML based models.

The collection, usage, and processing of investors’ personal data, as well as security measures, should comply with applicable laws.  Information about technical glitches and data breaches shall be communicated to SEBI and other relevant authorities, as per existing regulatory and legal framework, it said.

The paper proposed a regulatory-lite framework for the usage of AI/ML in the securities market for any purpose other than business operations that may directly impact their customers. The regulator has sought public comments on the discussion paper by July 11, 2025.

Also read: Tuhin Kanta Pandey brings a light touch to SEBI compared to his predecessor