POLITICS

Delhi Police Deploys AI Surveillance at Jantar Mantar Protests, Sparking Privacy Lawsuit

The capital’s police force has rolled out an AI‑driven monitoring network at the Jantar Mantar protest camp, using facial‑recognition cameras, body‑worn devices and mobile surveillance vans. A public‑interest litigation argues the continuous biometric tracking violates constitutional privacy and the right to peaceful assembly.

By Open Vaartha Desk ·

TL;DR

Delhi Police’s AI surveillance at Jantar Mantar protests pits public‑order aims against constitutional privacy rights, prompting a high‑court challenge.

Key points

<p>The Jantar Mantar site in Delhi has become the focal point of a series of anti‑government protests that began in early June 2026. In response, the Delhi Police announced the deployment of an extensive AI‑driven surveillance ecosystem to monitor the encampment. Officials say the technology is meant to deter violence and locate individuals with pending criminal cases. Critics, however, argue that the measures amount to continuous biometric tracking of peaceful demonstrators, raising questions about the compatibility of such surveillance with constitutional guarantees of privacy and free assembly.</p><p>The surveillance suite combines AI‑enabled CCTV cameras, facial recognition software, automatic number‑plate recognition, crowd‑density analytics and hand‑gesture detection. Body‑worn cameras and mobile surveillance vans equipped with eight‑camera 360° rigs—known as the Ikshana vehicle—feed live video to the Integrated Command, Control, Communication and Computer Centre (C4I). Officers also use Meta smart glasses that overlay facial matches in real time. Together, these tools create a layered monitoring network that can flag a suspect the moment a face or vehicle appears in the camera’s field of view.</p><p>The Ikshana mobile unit, first introduced ahead of the 2023 G20 Summit, is a purpose‑built van housing eight high‑resolution cameras that swivel to achieve a full 360° panorama. Its onboard AI processor runs facial‑recognition algorithms against police databases, producing instant alerts when a similarity score exceeds the 80 % threshold cited in a 2022 Right‑to‑Information reply. A human operator then validates the match before any further action. The vehicle also serves as a mobile command centre, allowing officers on the ground to coordinate responses without returning to a fixed control room.</p><p>According to the police briefing, Delhi’s AI surveillance network now comprises roughly 25,000 CCTV installations, of which more than 15,000 were already operational before the protest. An additional 2,100 AI‑enhanced cameras have been installed in the weeks following the demonstrations, linking the city’s Safe City Project, the Delhi Metro Rail Corporation and Indian Railways to a single monitoring platform. This integration enables cross‑agency tracking of persons of interest across public transport hubs, streets and the protest site itself, effectively expanding the city’s biometric footprint.</p><p>A public interest litigation lodged in the Delhi High Court alleges that surveillance has been continuous since 20 June 2026, capturing protesters while they demonstrate, eat, rest and receive medical aid. The petition further claims police threatened to forward images and videos to families, schools and potential employers, thereby weaponising biometric data against civilians. The filing points to the Supreme Court’s 2017 Puttaswamy judgment, which enshrines privacy as a fundamental right, and to the Malak Singh precedent that bars arbitrary police monitoring of lawful assemblies.</p><p>Delhi Police maintains that photography and videography are routine law‑enforcement tools, and that AI augmentation merely enhances the ability to maintain order and identify known offenders. In its statement, the force emphasized that the system is not intended for indiscriminate public monitoring but for targeted identification of individuals with outstanding warrants. Officials also argue that the technology has helped avert clashes in previous protests by providing early warnings of crowd surges and potential violent flashpoints, thereby protecting both demonstrators and by‑standers.</p><p>Despite the police’s assurances, India currently lacks a dedicated statute governing the use of facial‑recognition systems in public spaces. The Internet Freedom Foundation has highlighted the absence of a Privacy Impact Assessment, limited transparency about data retention, and unclear safeguards for biometric information. The Puttaswamy test—legal basis, legitimate objective, proportionality and least restrictive means—has yet to be applied to the Delhi Police’s AI deployment, leaving a legal vacuum that civil‑rights groups warn could chill the exercise of peaceful assembly.</p><p>Globally, approaches to live facial recognition vary widely. The European Union largely bans the practice for law‑enforcement, while Sweden permits it only for serious crimes under strict oversight. The United States leaves regulation to individual states, and the United Kingdom is expanding its use despite privacy concerns. China operates the most extensive system, integrating facial data across public and private domains. India’s nascent framework places it between the permissive U.S. model and the restrictive EU stance, and the outcome of the current PIL could set a precedent for how biometric surveillance is balanced against democratic freedoms.</p>