Facial recognition: A technology of promise and peril under Supreme Court scrutiny

The Supreme Court on Tuesday (August 18, 2026) agreed to examine the “proportionality” of the use of facial recognition systems (FRS) deployed during the students’ protest in the national capital last month. This follows an affidavit filed by the Delhi Police admitting to the use of the technology at the protest site.

In its affidavit, the police maintained that the FRS does not automatically capture, create, generate, or maintain profiles of every individual present at the protest site. It further said that the system is not deployed for indiscriminate surveillance or collection of personal information of peaceful protesters unless they have a previous criminal record. The police described the use of the technology as a “legitimate, bona fide, and proportionate policing measure”.

However, the very nature of FRS and its potential, if left unregulated, raises serious concerns about its implications for public behaviour and privacy in India. The Supreme Court’s decision to scrutinise the proportionality of its use marks a crucial moment for the legal framework governing such technologies.

How does facial recognition work?

While the use of AI-powered machine learning to monitor a mass gathering may seem novel, facial recognition has become a part of everyday life across the world. From unlocking mobile phones to contactless air travel under the Digi Yatra initiative, Indians are increasingly exposed to various degrees of FRS. The technology is broadly used in two ways: verification and identification.

Verification is the more common application, confirming a person’s identity against previously provided data, such as in phone security systems. Identification, on the other hand, scans distinguishing facial features and matches them against a database to determine a person’s identity without requiring them to disclose it. While verification is typically voluntary, identification may occur without the consent of those being monitored.

When an image is entered into an FRS for identification, the programme converts facial features into numerical data—distances between eyes, nose length and shape, wrinkle patterns, and other distinguishing characteristics. This data is then compared with a pre-existing database of faces to find a match. However, because the system depends on a database that may be flawed or manipulated, and because AI systems are prone to errors or “hallucinations”, FRS identifications are not always accurate. This lack of accuracy is a major technical and ethical concern, as biased or faulty systems can lead to false detentions and wrongful accusations.

How is FRS being used in India?

In its affidavit, the Delhi Police reiterated that FRS does not automatically maintain profiles of individuals or collect personal information indiscriminately. In March 2020, the Union Home Ministry approved the implementation of the Automatic Facial Recognition System (AFRS) by the National Crime Records Bureau (NCRB). The system was intended to use police records and would be accessible only to law enforcement agencies, as stated by then Union Minister of State for Home Affairs G. Kishan Reddy in a written reply in the Rajya Sabha.

Law enforcement agencies also have access to First Information Reports, chargesheets, and investigation reports from police stations across the country through the Crime and Criminal Tracking Network and Systems (CCTNS). This network includes past and current crime records, including personal details of accused persons, charge-sheeted individuals, convicts, habitual offenders, and proclaimed offenders. CCTNS is also linked to the National Intelligence Grid (NATGRID), a secure platform intended for police and investigating agencies.

While the government and police argue that FRS is a necessary tool for law enforcement, civil liberties advocates warn that unregulated use of such surveillance technology could erode fundamental rights. The Supreme Court’s examination of the “proportionality” of its use will be a key test of how India balances security concerns with individual privacy and constitutional freedoms. The outcome of this case could set a precedent for the deployment of facial recognition across the country.