Delhi Police's Defense: 'We Only Flag Criminal Records'

Delhi Police has told the Supreme Court that its use of facial recognition technology (FRT) during protests linked to the Chief Justice's Protection (CJP) matter was narrowly targeted. According to the force, cameras deployed at these gatherings were used solely to identify individuals with existing criminal records, not to build a database of every attendee.

On paper, this sounds like a reasonable, proportionate use of surveillance technology: scan the crowd, cross-reference against a watchlist, flag known offenders. But the explanation raises an obvious technical and legal question that the police submission doesn't fully answer. How does a facial recognition system identify someone with a criminal record without first scanning and processing the face of everyone in the frame?

The Gap Between Stated Purpose and Actual Deployment

Facial recognition doesn't work by magically skipping over law-abiding citizens. Every system of this kind operates the same way: it captures facial data from everyone present, converts those images into biometric templates, and then runs each one against a reference database. The "criminal records only" framing describes the output of the system, not its input. Every person at a CJP protest, regardless of their history or intent, would have had their face captured and processed as part of that matching pipeline.

This is not a new pattern in how facial recognition has been justified in India. As MediaNama has reported in the past, similar mobile surveillance units have shown up at other protest sites, including gatherings tied to the NEET paper-leak controversy, where crowds were monitored using comparable technology. That earlier episode is part of a broader pattern examined in India's facial recognition surveillance has no legal limits, which lays out how these systems have been deployed at public gatherings without a dedicated legal framework governing their use, retention, or oversight.

What's notable in the Delhi Police submission to the Supreme Court is the framing itself: it centers on intent (identifying criminals) rather than method (scanning everyone). That distinction matters enormously for privacy law, because consent and proportionality are usually judged by what data is collected, not just what it's later used for.

Why This Matters Beyond One Court Filing

India currently has no dedicated statute specifically regulating facial recognition technology. Police departments and other government agencies have rolled out FRT systems for crowd monitoring, criminal identification, and even traffic enforcement, largely under general policing powers rather than purpose-built legislation. That regulatory gap is precisely why the same concerns keep resurfacing: protest surveillance, crowd scanning, and biometric data collection without meaningful legal limits have become a recurring feature of how Indian law enforcement approaches public gatherings.

The Supreme Court's handling of this case could set an important precedent. If the court accepts the "criminal records only" justification without probing the underlying scanning process, it effectively endorses mass biometric collection at protests as long as police can point to a narrow downstream use. If, instead, the court pushes for clarity on retention periods, consent mechanisms, and oversight of the scanning process itself, it could push India closer to the kind of accountability frameworks seen in jurisdictions with dedicated biometric privacy laws.

What This Means For You

If you attend public demonstrations, rallies, or protests in India, it's worth assuming that facial recognition cameras may be present, regardless of the stated purpose of the deployment. The distinction between "we only flag criminals" and "we scan everyone to find criminals" isn't just semantic. It determines whether your biometric data was captured, stored, or shared, even if you were never suspected of wrongdoing.

This case is also a reminder that surveillance infrastructure tends to expand quietly. A system justified for one narrow purpose, identifying flagged individuals, can just as easily be repurposed for broader monitoring of attendees, organizers, or even journalists covering an event, especially in the absence of clear legal boundaries.

Actionable Takeaways

  • Stay informed about ongoing court proceedings regarding facial recognition, since rulings in cases like this one will shape how much oversight police face going forward.
  • If privacy at public gatherings concerns you, consider that facial recognition captures data regardless of whether you're a target, making blanket assumptions about "only criminals are affected" worth questioning.
  • Support and follow organizations pushing for dedicated facial recognition legislation in India, since the current absence of clear rules is what allows these gray-area deployments to continue.
  • Keep an eye on how this Supreme Court matter develops, as its outcome could either legitimize current practices or force police departments to disclose more about how their systems actually operate.

The core tension in this story isn't whether Delhi Police has a legitimate interest in identifying people with criminal histories. It's whether achieving that goal requires scanning everyone else in the process, and whether anyone outside the police department gets to say no.