Supreme Court Declines Plea Seeking Regulation Of AI Use By Central And State Governments; Asks Centre To Consider Representation

The Supreme Court today declined to entertain a plea seeking judicial guidelines to regulate the use of Artificial Intelligence (AI) by the Union and State Governments, including safeguards for high-risk AI systems used in areas such as welfare, policing, surveillance and content moderation.

A bench of Chief Justice of India Surya Kant and Justices Joymalya Bagchi and V Mohana observed that the issue falls within the policy domain and asked the Centre to consider a representation already made by the petitioner on the issues.

The petition sought a declaration that unregulated deployment of AI in governance and surveillance without a statutory framework, mandatory human oversight and a guaranteed right to explanation violates Articles 14, 19(1)(a), 19(1)(g) and 21 of the Constitution.

Alternatively, it sought a declaration that any State AI system producing civil consequences must satisfy minimum constitutional safeguards of legality, transparency, non-arbitrariness, proportionality, human oversight and effective remedy.

During the hearing today, petitioner-in-person NK Goswami submitted that he was pressing prayer D of the petition, which seeks a direction to the Union to file an affidavit disclosing all existing and proposed high-risk AI systems used by Union Ministries and agencies.

The CJI observed that the petition was very comprehensive and said that it should be sent as a representation to the authorities. “Two reasons it cannot be addressed by us. One is it is a highly technical issue. We are not the experts. And this is a policy domain,” he said.

The Court also noted that the use of AI by the Judiciary was already regulated. “As far as AI use in Judicial system is concerned, we have already well-regulated it. Our regulations are very comprehensive, you can go on the website and find it,” the CJI said.

The Court's noted that the petitioner had already made a representation to the Centre in February, comprehensively explaining the issues raised by the petitioner.

“We find that the issue is the formulation of appropriate policy and/or regulations for governing the use of AI and similar tools…We dispose of this writ petition at this stage without expressing any opinions on the merits, giving liberty to the respondents to consider the suggestions for the purpose of taking appropriate measures,” the Court ordered.

It permitted the petitioner to supplement the representation by sending a copy of the writ petition to the respondents.

The writ petition sought a wide framework governing the State's use of AI. It sought directions to the Union to constitute a High-Powered Expert Committee, with nominees of the Supreme Court, to frame binding guidelines on AI ethics, mandatory algorithmic impact assessments and bias audits, transparency in AI-based surveillance and content moderation, human-in-the-loop oversight for all high-risk governmental AI systems and data protection safeguards. The petitioner seeks that these guidelines be framed within three months and operate as law under Articles 141 and 142 of the Constitution until parliamentary legislation is enacted.

Pending such guidelines or a statutory framework, the petition sought an interim direction restraining the Union, its agencies and State Governments from deploying or substantially expanding new high-risk AI, Machine Learning or automated decision-making systems having civil consequences upon welfare entitlements, liberty, privacy, movement, livelihood, policing, surveillance or freedom of speech, unless specified constitutional safeguards are ensured.

The petition sought safeguards including public disclosure of the system's existence and purpose, a prior algorithmic impact assessment, an independent bias or accuracy audit wherever feasible, identification of a human officer responsible for final decision-making, notice and reasons before adverse action, meaningful human review, grievance redressal and restoration in cases of wrongful exclusion or other adverse consequences.

The petitioner also sought directions for a comprehensive algorithmic impact assessment and independent bias audit of all existing AI systems deployed in welfare allocation, predictive policing, facial recognition and content moderation. The petition sought that these assessments be made public within six weeks along with a complete inventory of such systems.

The petition also sought a direction to the Union Government to immediately take steps to enact comprehensive parliamentary legislation regulating the development and deployment of AI by the State, consistent with guidelines framed by the Supreme Court.

Case no. – W.P.(C) No. 837/2026

Case Title – Narendra Kumar Goswami v. Union of India