The Supreme Court of India on Thursday directed the Union government to evaluate proposals seeking binding ethical standards, transparency, and oversight mechanisms for artificial intelligence tools deployed across state surveillance, policing, and content moderation.
Disposing of a public interest litigation without ruling on its merits, a bench comprising Chief Justice Surya Kant, Justice Joymalya Bagchi, and Justice V. Mohana instructed relevant authorities to treat the petition as a formal representation and consider appropriate policy measures. The court also granted the petitioner liberty to submit supplementary material to the government.
Policy Domain Governance
Addressing the petitioner during the hearing, the bench remarked that regulating artificial intelligence is a complex technical matter belonging strictly to the policy domain. The judges noted that the court lacks specialized technical expertise to formulate operational frameworks for advanced technology, emphasizing that policy formulation rests with executive authorities.
The direction came on a petition filed by advocate Narendra Kumar Goswami, who had previously submitted a representation to the government in February regarding artificial intelligence governance. Goswami urged the court to compel the central government to disclose existing and proposed high-risk artificial intelligence applications utilized by federal ministries and state agencies in law enforcement, public welfare distribution, surveillance, and content regulation.
Key Demands For Regulation
The petition called for the establishment of a high-powered expert committee to design mandatory ethical guidelines, data protection safeguards, and algorithmic impact frameworks. It argued that all state-run artificial intelligence systems operating in high-risk sectors must incorporate mandatory human oversight to prevent automated errors and bias.
Additionally, the plea urged the court to direct the government to conduct independent bias audits and comprehensive impact assessments on all operational systems within six weeks and make the findings public. These systems include tools used for predictive policing, facial recognition technology, content moderation, and welfare allocation.
The petitioner further requested directions for the government to enact comprehensive parliamentary legislation to regulate state development and deployment of artificial intelligence in accordance with judicial oversight principles
