On July 23, 2026, journalist Ila Kazmi posted a reel showing a Delhi Police van at Jantar Mantar employing facial recognition technology to match protesters’ faces in real time. Within a day the video disappeared from India, and on September 9, 2026 her Instagram account was blocked. Meta said it complied with a legal request. The matter reached the Delhi High Court on October 9, 2026, which ordered the central government, Delhi Police and Meta to disclose the orders behind both blocks.
Key Developments
- July 23 – Reel showing police use of facial‑recognition at a major protest site goes viral.
- July 24 – Video becomes unavailable in India.
- September 9 – Instagram disables Kazmi’s account, citing a legal notice.
- October 9 – Delhi High Court directs authorities to produce the underlying legal orders.
- Indian Express investigation finds at least 25 individuals flagged by the tool were already in jail, highlighting algorithmic errors.
Important Facts
The surveillance tool is known to produce false matches, yet it was deployed during a politically sensitive protest. The blocking of the explanatory video limits public scrutiny, raising concerns about transparency and accountability in the digital public square. The court’s order seeks to uncover the legal basis for both the police surveillance and the social‑media restriction.
Exam Relevance
Understanding this episode helps aspirants in multiple GS papers:
- GS2 (Polity): Examines the balance between law‑enforcement powers, privacy rights, and judicial oversight.
- GS3 (Technology & Economy): Highlights challenges of AI‑driven surveillance, data protection, and the regulatory framework for tech platforms.
- GS4 (Ethics): Raises ethical questions about state surveillance, freedom of expression, and the role of digital media in a democracy.
Way Forward
1. Legislative clarity: Parliament should enact a comprehensive data‑protection law that defines permissible use of facial‑recognition by public agencies.
2. Judicial oversight: Courts must ensure that any restriction on online content is backed by transparent, time‑bound orders.
3. Technical safeguards: Police should adopt error‑rate reporting and human‑in‑the‑loop verification before acting on AI matches.
4. Civic awareness: Media literacy programmes can help citizens understand how surveillance tools work and their rights under the law.