Editorial-22/08/2026
Saving faces: On protests, biometric surveillance
The use of AI-powered facial recognition and other biometric surveillance tools at public protests in India has escalated from speculative fear to documented reality, raising urgent questions about the legality, proportionality, and democratic cost of such practices. Recent student-led demonstrations over the NEET-UG exam paper leak saw Delhi Police deploy a sophisticated surveillance apparatus—including AI-enabled cameras, smart glasses, drones, and mobile command vans—capable of scanning and matching protesters’ faces in real time against criminal databases. Yet, there is no specific law authorising such mass, indiscriminate biometric processing of peaceful assemblies, and existing data-protection frameworks offer broad exemptions to the state, leaving citizens’ fundamental rights to privacy, free speech, and assembly dangerously exposed.
The surveillance architecture at recent protests
During the Cockroach Janata Party (CJP) protests at Jantar Mantar in July–August 2026, Delhi Police acknowledged in court filings that it used an AI-powered facial recognition system, integrated with a network of CCTV cameras, smart glasses worn by personnel, and a mobile surveillance van linked to private contractors. The system, reportedly named “Ikshana,” enabled real-time facial scanning of individuals within camera range, with footage potentially matched against databases including the National Crime Records Bureau’s Automated Facial Recognition System (AFRS). Petitioners before the Delhi High Court and Supreme Court allege that the technology was deployed indiscriminately, capturing the biometric data of thousands of peaceful protesters—many of them students—without consent, clear purpose limitation, or transparency about retention or access.
The scale and intrusiveness of this deployment are unprecedented in India’s recent protest history. Unlike targeted surveillance of specific suspects, this system processed every face in a crowd, converting each into a searchable biometric template. The petitioners further claim that police personnel threatened to share footage of young protesters with their families and institutions, adding a layer of coercive intimidation to the technological overreach.
Legal vacuum and constitutional concerns
India currently has no statute that expressly authorises, regulates, or limits the use of facial recognition technology (FRT) by police at public gatherings. The Criminal Procedure (Identification) Act, 2022, often cited by law enforcement, applies only to persons arrested, convicted, or otherwise brought within the criminal justice process—it does not cover citizens exercising their fundamental right to assemble peaceably under Article 19(1)(b) of the Constitution. Similarly, the Digital Personal Data Protection Act, 2023, which has not yet come into force for data-processing obligations, includes broad exemptions for state processing in the interests of sovereignty, security, and public order, effectively leaving government surveillance unchecked.
This legal vacuum means that the deployment of FRT at protests fails the first prong of the K.S. Puttaswamy privacy test: legality. Under Puttaswamy, any state action infringing privacy must be authorised by law, serve a legitimate aim, and be proportionate. Without a law, the question of proportionality becomes moot—but even if one were imagined, mass biometric scanning of peaceful assemblies would struggle to meet the necessity and least-restrictive-means criteria. Traditional policing methods—such as visual identification of specific suspects or targeted monitoring based on credible threats—are less intrusive and equally effective for maintaining public order.
The indiscriminate nature of FRT also raises a second constitutional issue: the chilling effect on free speech and assembly. When citizens know that attending a protest could result in their biometric data being logged, stored, and potentially cross-referenced with criminal databases for unknown future use, they may self-censor or avoid participation altogether. This “awareness of being watched” reshapes conduct and denies speech before it is even uttered—a harm recognised by courts globally, including in the UK’s Bridges case, which struck down live facial recognition use due to the absence of a clear legal framework.
Comparative lessons and the path forward
India is not alone in grappling with the democratic risks of protest surveillance. In 2020, the England and Wales Court of Appeal held that police use of live facial recognition violated privacy rights because the legal framework left too much discretion to individual officers, with no clear policy on who could be scanned or where. The court also noted failures to assess bias or error rates, particularly for marginalised groups. In the United States, senators and civil rights groups have similarly condemned the use of mobile facial recognition by Immigration and Customs Enforcement (ICE), warning that it threatens First Amendment rights and disproportionately harms communities of colour.
These comparative signals underscore a structural lesson: it is not the technology itself, but the absence of a clear, public, and judicially reviewable framework that renders its use unlawful. States cannot outsource the boundaries of fundamental-rights-infringing powers to unpublished internal orders or private contractors.
In India, the Supreme Court’s agreement to hear petitions challenging FRT use at protests offers a critical opportunity to establish doctrinal guardrails before executive practice becomes irreversible. Petitioners are not asking for a permanent ban on the technology, but for a declaration that its deployment at peaceful assemblies is unconstitutional until Parliament legislates. This sequencing is correct: courts should not design surveillance frameworks from the bench, but they can and must insist that Article 21’s guarantee of “procedure established by law” be honoured before such powers are exercised.
What a regulatory framework must include
Any future legislation governing FRT use at protests must be narrow, transparent, and rights-respecting. Key elements should include:
- Purpose limitation: FRT should be deployable only for specific, credible threats to public order, not for general monitoring of assemblies.
- Prior authorisation: Deployment should require prior judicial or independent oversight, not unilateral police discretion.
- Data minimisation and retention caps: Non-matching data should be deleted immediately; matching data should be retained only for a strictly defined period and purpose.
- Accuracy and bias audits: Regular testing for error rates, particularly across gender, race, and age groups, with public reporting.
- Transparency and accountability: Clear public policies on when, where, and how FRT is used, with avenues for redress and oversight.
Until such a framework exists, the continued use of FRT at protests remains constitutionally suspect and democratically corrosive.
The broader context: shutdowns, opacity, and the right to dissent
The biometric surveillance of protesters is part of a broader infrastructure of control that includes internet shutdowns, signal jamming, and opaque data-sharing with private contractors. During the CJP protests, mobile internet was suspended in micro-geofenced zones around Jantar Mantar, severing demonstrators’ ability to document state conduct and coordinate in real time. While the Supreme Court’s 2020 Anuradha Bhasin judgment held that internet access is constitutionally protected and suspensions must meet strict necessity and proportionality tests, in practice, orders are often withheld or published retroactively, shielding administrative decisions from timely review.
Together, these tactics render citizen dissent invisible to the outside world while hyper-legible to the state—a strategy that may succeed in suppressing short-term disruption but risks long-term democratic erosion. As the Supreme Court considers the constitutional challenge to FRT, it has the opportunity to reaffirm that the right to assemble is not a privilege to be monitored, but a freedom to be protected.
The question before the Court is not whether technology can be used to police protests, but whether the state can do so without law, oversight, or regard for the chilling effect on democracy itself. The answer will shape what it means to assemble in public in India for generations to come.
Download Pdf
