Supreme Court Asks Centre To Consider Representation Against Unregulated Use Of Artificial Intelligence By State

The Supreme Court disposed of a Public Interest Litigation filed under Article 32 seeking constitutional safeguards and a regulatory framework against the unregulated deployment of Artificial Intelligence, machine learning, and automated decision-making systems by the State.

The Court asked the Union government to treat the petition and the petitioner's comprehensive presentation as a representation for taking appropriate policy measures, noting that the issues involved complex technical and policy domains outside judicial expertise.

The petitioner, an advocate appearing in person under Article 32 of the Constitution of India, instituted a Public Interest Litigation seeking a writ of mandamus and constitutional safeguards against the unregulated deployment of Artificial Intelligence (AI), Machine Learning (ML), and automated decision-making systems by the State.

The Bench of Chief Justice Surya Kant, Justice Joymalya Bagchi and Justice V Mohana ordered, "The Petitioner, who is a practising advocate, asserts that unregulated and non-transparent deployment of Artificial Intelligence and other assist technology, like machine learning, by the state and its instrumentalities in matters of deciding welfare scheme entitlements, predictive policing, and social media content regulation have a direct impact on civil liberties, privacy, dignity, and livelihood of people...We find that the issues raised, formulation of appropriate policy and/or regulations for governing the use of AI or similar tools, the petitioner has already submitted a comprehensive presentation on 12th February 2026...We, therefore, dispose of this petition at this stage, without expressing any opinion on merits...The respondents to consider suggestions for the purpose of taking appropriate measures...The petitioner shall be at liberty to supplement the representation by sending a copy of the writ petition to the respondents."

The petitioner, Advocate Narender Kumar Goswami, submitted that Artificial Intelligence is no longer merely an administrative aid or technological convenience.

Chief Justice Kant said, "Tell us what can we do? We can send it as a representation. It is a very comprehensive petition. We have two issues: one is highly technical issues as we are not the experts; secondly, policy domain...".

The Plea highlighted that the Digital Personal Data Protection Act, 2023, while dealing with certain aspects of digital personal data, is not a comprehensive rights-based AI governance statute.

It said that it does not create an enforceable constitutional regime for algorithmic impact assessment, independent bias audit, explainability, human review, demographic impact assessment, public disclosure of high-risk AI systems, or remedies against algorithmic exclusion and harm. Similarly, policy papers, advisories, discussion documents and ethical frameworks, however useful, do not constitute binding “law” within the meaning of Article 13 and cannot by themselves authorise State action infringing fundamental rights.

It was stated in the plea, "It has begun to enter the bloodstream of governance. Public-domain official materials, policy documents and supporting documents placed with the Petition show that AI/ML, automated decision-making, fraud-detection, beneficiary-verification, Aadhaar-linked authentication, data-driven de-duplication and similar algorithmic systems are being used, piloted or proposed in welfare-linked governance, including public distribution, health-benefit delivery, direct-benefit transfer, ration-distribution optimisation, beneficiary identification and fraud-prevention mechanisms. These systems, by their nature, are capable of generating automated flags, risk indicators, eligibility signals, fraud alerts or exclusionary outputs that may affect access to essential welfare benefits."

The Petition interlinked three constitutional concerns:

1. In the field of welfare allocation and public-benefit delivery, automated decision-making, fraud detection, risk scoring, Aadhaar-linked authentication, de-duplication and beneficiary-verification systems may affect the poor, Scheduled Castes, Scheduled Tribes, minorities, women, elderly persons, persons with disabilities, migrant workers and digitally excluded citizens most severely.

2. In the field of law enforcement and surveillance, facial recognition systems, biometric identification tools, predictive policing mechanisms and similar AI-assisted technologies raise serious questions of privacy, dignity, autonomy, movement, association and non-discrimination. The deployment of such technologies without a specific statutory framework, judicial or independent oversight, data-retention limits, accuracy disclosure, bias audit and grievance redressal fails the constitutional tests of legality, necessity and proportionality laid down by the in Justice K.S. Puttaswamy (Retd.) v. Union of India.

3. Thirdly, in the field of online speech and content regulation, AI-assisted or automated content moderation, takedown recommendation, cyber-flagging and speech-classification systems raise grave concerns under Article 19(1)(a). Any State-influenced or State-controlled system that suppresses, blocks, flags, downranks or recommends takedown of online speech without a transparent legal standard, without notice to the speaker, without a reasoned human order and without an effective appellate remedy carries the danger of algorithmic prior restraint and a chilling effect on journalism, satire, political criticism and democratic dissent.

The Plea prayed, "Direct the Union to file before this Hon’ble Court a complete affidavit disclosing all existing and proposed high-risk AI systems used by Union Ministries/agencies in welfare, policing, surveillance and content regulation, together with their legal basis, vendor/agency, purpose, data categories, safeguards, grievance mechanism and audit status."

The petitioner prayed for issuance of a writ of mandamus directing the Union of India to constitute a High-Powered Expert Committee to frame binding Vishaka-style guidelines for AI regulation in governance, pending parliamentary legislation.

It also prayed for an interim injunction restraining the Union and State governments from deploying or expanding new high-risk AI systems in welfare, surveillance, policing, or speech moderation without prior constitutional safeguards.

Cause Title: Narender Kumar Goswami v. Union of India & Ors. [W.P.(C) No. 837/2026]