On June 3, 2026, the Supreme Court of India issued a ‘draft’ through the Artificial Intelligence Commission.Regulations on the Use of Artificial Intelligence (AI) in Courts, 2026This draft document represents an important step toward integrating emerging technologies into law enforcement systems in a systematic and responsible manner.
The proposed framework reflects a thoughtful approach to the adoption of AI in courts, with provisions based on widely accepted principles such as human primacy, transparency, accountability, and accountability. These principles are consistent with global standards, including the UNESCO Guidelines for the Use of AI Systems in Courts and Tribunals (2025), and reinforce the need to ensure that technology remains an adjunct to, rather than a replacement for, judicial decision-making.
The draft regulation also outlines the extent to which the use of AI within the justice system is permitted and prohibited, and introduces safeguards such as auditing requirements and monitoring mechanisms. This demonstrates a clear intention to carefully balance innovation, especially in an area as sensitive as the delivery of justice.
Although this framework is a welcome development, certain aspects could benefit from further clarity to support consistent interpretation and effective implementation.
One key area relates to the use of risk-based terminology within the draft. Terms such as “significant risk of harm,” “high-risk applications,” and “risk profile of the associated task” are referenced across various regulations. However, in the absence of clear definitions or illustrative guidance, these terms are likely to be interpreted differently by different parties. Providing examples of how risk levels can be assessed across different use cases can help build a more unified understanding. Furthermore, linking such risk classifications to the permissible use cases identified in the draft law could enable a more structured approach to deploying AI systems in courtrooms.
Another area that requires clarification is the scope of audit requirements, particularly for technical audits. The draft proposal suggests a level of access to AI systems that is not always possible in practice, especially if courts rely on third-party or proprietary solutions. In these cases, access to the source code or underlying algorithms is typically restricted. Clarifying the circumstances in which such access is required and distinguishing between different deployment models will help align regulations with common industry practices.
Overall, the draft regulations represent a significant milestone in shaping the future of AI in the Indian judiciary. With targeted clarification and refinement, this framework could provide a solid foundation for responsible and effective AI implementation in courtrooms.
For more information, please see the submission attached below.
If you have any suggestions or questions, please contact Saumya Anand. sanand@nasscom.in or with Sudipto Banerjee sudipto@nasscom.inCopy it to policy@nasscom.in.
