The use of Facial Recognition Technology at sites of peaceful protest risks converting dissent into data. Critically examine in the light of India's privacy jurisprudence.

In K.S. Puttaswamy (2017), the Supreme Court held that any intrusion on privacy must have a public purpose, a rational link to that purpose, and be the least intrusive, proportionate means [1]. Rajya Sabha MP A.A. Rahim's challenge to police FRT and biometric surveillance of NEET-UG protesters at Jantar Mantar tests whether crowd surveillance meets this standard [2].

How FRT turns dissent into data

• Legality gap: India has no law made for FRT. The Criminal Procedure (Identification) Act, 2022 covers photographs only of convicts, arrested persons and detainees [3]. Scanning a whole crowd falls outside it.
• Retention: records are kept for 75 years and are deleted only after acquittal or release without trial [4]. PRS says this may fail the necessity and proportionality test [1].
• Burden on the citizen: a person must ask a nodal officer to delete a record [3]. They cannot do that for a record they cannot see.
• Chilling effect: NITI Aayog's Responsible AI for All paper says FRT puts privacy, equality, free speech and movement at risk [5]. Fear of being profiled weakens Article 19(1)(a)–(b).
• Function creep: NCRB can share and disseminate records [4]. CCTNS links about 16,600 police stations, so data can spread across the country [6].

The State's counterpoints

• Safeguards in the Rules: people held only under Sections 144/145/151 CrPC cannot be measured unless they are also charged with another offence [3].
• Burden of proof: MHA/NCRB call the claims "omnibus". They argue that a general allegation is no substitute for proof of even one unlawfully stored record [2].
• Legitimate aim: Puttaswamy allows restrictions that serve a public purpose [1], such as identifying violent people in a crowd.

Assessment: The State's points have some force. But it asks for "foundational facts" that only its own databases hold. Its safeguards also sit in Rules and NCRB SOPs that Parliament may never review [3], so they do not satisfy Puttaswamy's requirement of a law.

FRT can help policing, but crowd scanning without a law fails the first test of the privacy framework. As a way forward, Parliament should pass an FRT-specific law built on NITI Aayog's risk framework [5]. Records should be deleted automatically, as under the Juvenile Justice Act, 2015 [3], and NCRB's SOPs should be placed before Parliament. Together, these would protect peaceful dissent under Articles 19 and 21 and keep it from becoming a permanent police record.