The deployment of artificial intelligence in State surveillance, ranging from automated facial recognition to predictive policing analytics and communication-interception software, has outpaced the constitutional and statutory framework designed to govern it. This article critically examines the tension between the right to privacy, recognised as intrinsic to Article 21 of the Constitution of India in Justice K.S. Puttaswamy v. Union of India, and the State's invocation of national security to justify AI-enabled monitoring of citizens. It traces the statutory basis for interception and surveillance under the Indian Telegraph Act, 1885, the Information Technology Act, 2000, and the Digital Personal Data Protection Act, 2023, arguing that each contains broad executive exemptions that were framed before artificial intelligence made mass, continuous and predictive surveillance technically and economically feasible. The article evaluates the judiciary's proportionality doctrine as applied in Puttaswamy, the Aadhaar judgment, Anuradha Bhasin, and the Pegasus spyware proceedings in Manohar Lal Sharma v. Union of India, and finds that while the doctrine offers a sound analytical framework, its enforcement against opaque, AI-driven and classified surveillance systems remains weak. Drawing a brief comparison with the European Union's risk-based regulatory approach and the United States' judicial and statutory responses to algorithmic policing, the article identifies structural gaps, the absence of an independent authorisation mechanism, the lack of algorithmic transparency, and the non-existence of a dedicated surveillance reform law, that leave Article 21 vulnerable in the age of artificial intelligence. It concludes with recommendations for a purpose-specific surveillance statute, independent oversight, and algorithmic accountability standards suited to the Indian constitutional order.

Keywords

Artificial intelligence; surveillance; right to privacy; Article 21; national security; facial recognition; Puttaswamy; proportionality; Digital Personal Data Protection Act; predictive policing.