Campaign
1 link
Graph · Event
01 · In focus
The structured facts the source records about Internet Freedom Foundation legal notice to Delhi Police demanding halt to Automated Facial Recognition System (28 December 2019), the count of declared adjacencies in the corpus, and the federation map zoomed on this node and its neighbours.
event
↑4 declared connections
02 · Connections
Split by direction. Direct links are the ones Internet Freedom Foundation legal notice to Delhi Police demanding halt to Automated Facial Recognition System (28 December 2019)’s source record names; inferred backlinks are records elsewhere in the corpus that point at this entity. Some records appear in both because the corpus names them from both sides — those rows carry a note.
3 links
Links named in this entity's structured fields.
1 link
1 link
1 link
1 link
Other records that name this entity.
03 · Background
Body prose as it appears in movement-graph’s published markdown for this entity. Links to other corpus entities resolve to their graph page; links to deeper repo paths are kept as text so the page does not invent a route.
On 28 December 2019 the Internet Freedom Foundation (IFF) sent a legal notice to the Secretary, Ministry of Home Affairs and the Commissioner of Police, Delhi demanding the complete halt and recall of Delhi Police's Automated Facial Recognition System (AFRS) — a system IFF called "an illegal act of mass surveillance." The notice, authored by IFF Policy Counsel Anushka Jain with input from researchers Vidushi Marda and Srinivas Kodali, was the first formal legal petition directed specifically at Delhi Police's expanding FRT use and the earliest anchor in the litigation trajectory that eventually produced the Project Panoptic tracker (November 2020) and the S.Q. Masood Telangana High Court public-interest litigation (January 2022). It predated the Project Panoptic tracker launch by nearly a year and was filed as direct legal pushback at a deployment moment IFF judged to represent both unchecked function creep from the system's authorised scope and a structural threat to the constitutional rights of protesters and ordinary Delhiites.
Delhi Police's Automated Facial Recognition System was sourced from M/s Innefu Labs Private Limited in March 2018 under the direction of the Delhi High Court in Sadhan Haldar v The State NCT of Delhi (WP(Crl) 1560/2017) — a 2017 petition seeking help to locate and reunite missing children with their families. The court order authorised the Delhi Police to obtain facial-recognition technology specifically for the purpose of tracking missing children. However, by the time IFF filed its legal notice, the police had expanded the system's use well beyond that authorised scope. In a Right to Information response IFF obtained in February 2020, Delhi Police confirmed that the technology is being used to match similar faces for police investigation — a function the original court order did not authorise and that IFF characterised as a textbook instance of function creep.
The immediate context for the December 2019 notice was Delhi Police's reported deployment of the AFRS at Prime Minister Modi's rally at Ramlila Maidan, screening attendees against facial data collected from protest videos to identify law-and-order suspects. The system was also deployed during the protests against the Citizenship Amendment Act that ran through late 2019 and into early 2020, with Delhi Police commissioner SN Shrivastava subsequently stating that 137 of 231 riot-related arrests were made with facial-recognition assistance following the February 2020 northeast Delhi communal violence. Minister of Home Affairs Amit Shah separately stated over 1,900 faces had been identified through the system in connection with those events.
The legal notice grounded its demand on three interlocking arguments. First, the accuracy record the Delhi Police had publicly disclosed was disqualifyingly low: the system operated at an accuracy rate of 2 per cent in 2018 and less than 1 per cent in 2019 — a figure the Ministry of Women and Child Development had independently noted as insufficient to distinguish between boys and girls. Deploying a system with less than 1 per cent accuracy in enforcement contexts carried a direct risk of coercive action — including arrest — against innocent citizens identified by false positives.
Second, the deployment pattern IFF documented violated the proportionality test the Indian Supreme Court's landmark judgment in Justice K.S. Puttaswamy v. Union of India (2017) applies to state actions that engage the right to privacy. The Puttaswamy proportionality standard requires that any state intrusion on privacy have a lawful purpose, be limited to what is necessary for that purpose, and be subject to procedural safeguards. The AFRS had neither a statutory basis (the Sadhan Haldar court order covered missing children, not general investigations or crowd surveillance at political events) nor any independent regulatory oversight, and the surveillance of protest crowds had no discernible connection to its statutory authorisation.
Third, IFF argued the deployment structure — scanning crowds at public events and building databases from footage of political protests — impaired citizens' constitutional rights to assembly, expression, and political participation. The system was not merely collecting biometric data; it was using the data to screen political-event attendees against protest footage, a deployment architecture that chills lawful protest participation by creating the realistic possibility that attendance at a demonstration will result in police identification and follow-up scrutiny.
Delhi Police's initial response to the legal notice was to assert that the AFRS incorporated industry-standard checks and balances and to deny that the system engaged in racial or religious profiling, stating that the database it matched against targeted terror suspects and individuals with criminal records rather than ordinary civilians. The response did not engage with the specific function-creep, accuracy, or proportionality arguments IFF had raised.
Following the legal notice, IFF pursued the transparency channel in parallel. On 24 January 2020 Anushka Jain filed a Right to Information request with Delhi Police (reference DEPOL/R/E/20/00519) covering eleven areas: legal authorisation for FRT deployment, prior privacy assessments, the purposes for which the technology was being used, whether it monitored protests or rallies, expenditure and procurement documentation, databases consulted, and accuracy rates. The February 2020 response confirmed the Sadhan Haldar court order as the claimed legal framework but simultaneously admitted the function creep IFF's legal notice had alleged: the police stated the technology was in use to "match similar faces for police investigation" — a use the original court order had not authorised.
On 23 October 2020 IFF filed a second batch of RTIs with Delhi Police (DEPOL/R/E/20/07128), Kolkata Police (initially filed 30 July 2020), and Telangana State Technology Services. Delhi Police's reply on 25 November 2020 refused to provide any information about the FRT system under Section 8(1)(d) of the RTI Act, invoking the commercial-confidence and trade-secrets exemption on behalf of the vendor Innefu Labs Private Limited. The trade-secrets refusal drew criticism from technology lawyers who noted that Section 8(1)(d) was being applied to protect a private vendor's commercial interests rather than any legitimate national-security ground. IFF filed First and Second Appeals through the established RTI appellate channel; at a Second Appeal hearing on 29 June 2022 the Central Information Commission directed Delhi Police to revise their responses in compliance with the RTI Act. Subsequent RTI responses following the CIC direction eventually disclosed that Delhi Police treats a face-similarity score above 80 per cent as a positive identification result for arrest purposes — a threshold Anushka Jain framed as treating 80 per cent probability as certainty and leaving a 20 per cent margin of wrongful-arrest error operative in enforcement decisions.
The 28 December 2019 legal notice is the corpus's first event anchored on a formal legal demand directed at Delhi Police's facial-recognition deployment and IFF's first documented strategic-litigation act in the surveillance-accountability space. It is structurally distinct from the Project Panoptic tracker launch on 27 November 2020 — which launched the public database, the public moratorium petition, and the citizen-reporting feature — and represents the earlier adversarial legal phase of IFF's work that preceded the tracker. The corpus previously had South Asia represented by a single event anchored on the tracker launch rather than on legal action; this notice entry closes the pre-tracker strategic-litigation slot.
Within the wider Project Panoptic campaign arc, the 28 December 2019 legal notice is the earliest documented legal act that the campaign's subsequent litigation track — the January 2020 and October 2020 RTIs, the CIC second appeal, and ultimately the January 2022 Telangana High Court PIL brought by S.Q. Masood with IFF legal-counsel support — builds on. The campaign's core legal arguments at each subsequent stage (no statutory basis, proportionality failure under Puttaswamy, function creep, wrongful-arrest risk from low accuracy) trace directly to the grounds IFF first articulated here. The notice also marks the first time IFF publicly framed Delhi Police's FRT deployment as a constitutional violation, a framing that carried forward unchanged into the Project Panoptic public petition, the #BanTheScan Hyderabad sub-campaign, and the Masood PIL in 2022.
04 · Sources
8 sources listed from the pinned corpus. Links are shown only when the source URL is a valid HTTP(S) address.
Project Panoptic case study on IFF's legal notice to Delhi Police — primary source for the notice date (28 December 2019), the addressees (Secretary Ministry of Home Affairs and Commissioner of Police Delhi), the demand for complete halt and recall of the AFRS, the Ramlila Maidan deployment context, the vendor identification as Innefu Labs Private Limited, the accuracy figures (2% in 2018, less than 1% in 2019), and the sequence of IFF actions through 2020
IFF blog post announcing the 28 December 2019 legal notice — primary source for IFF framing the AFRS deployment as an illegal act of mass surveillance, the notice demand for complete halt and recall, and IFF Policy Counsel Anushka Jain as the notice author
Inc42 technology-press coverage dated 30 December 2019 — independent secondary source for the 28 December 2019 notice date, the named addressees, IFF characterisation of the notice as targeting illegal mass surveillance, and the Delhi Police response denying racial or religious profiling and stating the system targeted terror suspects and individuals with criminal records
Project Panoptic running Delhi Police case study — primary source for the full chronology of IFF Delhi Police FRT actions including the legal notice, the January 24 2020 RTI filing, the February 2020 RTI response confirming function creep (FRT used to match similar faces for police investigations rather than only to locate missing children), and Amit Shah naming 1900 faces identified and 137 arrests made by FRT during the 2020 northeast Delhi riots
Project Panoptic record of IFF RTI to Delhi Police filed 24 January 2020 (reference DEPOL/R/E/20/00519, filed by Anushka Jain) — primary source for the eleven areas of inquiry the RTI covered (legal authorisation, protest surveillance, databases, accuracy rates, procurement, prior legal opinions and privacy assessments), and the February 24 2020 response date
Project Panoptic RTI update case study — primary source for IFF second batch of RTIs filed 23 October 2020 to Delhi Police (reference DEPOL/R/E/20/07128), Kolkata Police (filed 30 July 2020), and Telangana State Technology Services (filed 23 October 2020); Delhi Police reply dated 25 November 2020 denying all questions under Section 8(1)(d) of the RTI Act (commercial confidence, trade secrets, intellectual property) on behalf of vendor Innefu Labs Private Limited
IFF report on Central Information Commission Second Appeal heard 29 June 2022 — primary source for the CIC directing Delhi Police to respond afresh to IFF RTI applications in compliance with the RTI Act, the procedural outcome of the transparency-blocking trade-secrets refusal
MediaNama August 2022 coverage of Delhi Police RTI responses following the CIC direction — independent technology-press secondary source for Delhi Police eventually disclosing that a face similarity check above 80 percent is treated as a positive identification result for arrest purposes, and for IFF Associate Counsel Anushka Jain framing that 80% similarity treated as 100% leaves a 20% margin of error constituting wrongful-arrest risk
Source: entities/events/event-iff-delhi-police-frt-petition-2020.md — movement-graph pin 5d136ad.