Citywide CCTV Rollout and Delhi Police Facial Recognition
This overview was generated with AI from the public sources listed below and checked against them before publication. It has not had a complete human fact-check — treat it as a starting point and follow the sources.
8 sourced claims on this page, none individually verified yet — turn on “Sources in the text” to see each one.
How this atlas is made · What “last verified” means · What “source-checked” means · The full record behind this page
Delhi Police treat an 80 percent similarity score as a positive match. Their own court filings put the system's accuracy at 2 per cent in 2018 and under 1 per cent in 2019.
what does this mean?
The strongest source is the operator's or the city's own reporting — an official report, dataset or press release. It grades who did the looking — not how well the project went. All four grades →Part of 1001 Smart Cities — 1,167 documented projects · 🏙️ 9 more in New Delhi · 🛡️ Safety & security guide · 📡 Data & IoT projects
Delhi's Public Works Department has installed roughly 280,000 street CCTV cameras since 2018, sited largely where resident welfare and market associations asked for them; in March 2025 the government ordered a physical audit after finding 263,000 recorded as installed, to check how many exist and still work. Separately, Delhi Police runs a facial recognition system procured in 2018 under a High Court order to trace missing children, later used to identify people after the 2020 north-east Delhi riots and at protest sites. Police treat an 80 percent similarity score as a positive match, and no Indian law specifically governs the technology.
📊 Impact
The Union Home Minister told Parliament in March 2020 that 1,922 faces had been identified with facial recognition after the north-east Delhi riots, matched against driving-licence and voter-ID data. In court filings Delhi Police had earlier put the system's accuracy at 2 per cent (2018) and under 1 per cent (2019).
🎓 Lesson
A system bought to find missing children was pointed at a protest, and no law had to change for that to happen — India still has none written for the technology. What moved instead was the threshold: Delhi Police decided that an 80 per cent similarity score counts as a match, on a system whose accuracy their own court filings put at 2 per cent and then under 1 per cent. Where the cameras go is settled the same way, by which resident and market associations asked for them, which is why the government had to order a physical audit to find out how many of the cameras on its books exist and still work.
Sources
Show 3 more sourcesShow fewer
📅 How old is the evidence?
Two clocks, kept apart: when this entry’s sources were published, and what has happened to its record in the atlas. Documented span: since 2018. 7 of 8 sources carry a publication date, from 15 August 2019 to 24 July 2026.
When the sources were published
- Comparitech · pressComparitech — The World's Most Surveilled Cities (research)
- The Week · pressThe Week — Facial recognition identified 1,900 Delhi rioters: how does the technology work
- Outlook India · pressOutlook India — Delhi Police's 80 per cent match threshold raises concern, finds RTI
- Business Standard · pressBusiness Standard — Delhi govt to conduct audit of over 260,000 CCTV cameras
- The Week · pressPTI via The Week — Delhi PWD street CCTV programme (14 May 2025)
- The News Minute · pressThe News Minute — Explained: Delhi Police's AI surveillance network and why it's under scrutiny
- The Tribune · pressThe Tribune — Facial recognition and AI surveillance vehicles deployed at Jantar Mantar protest site
Show your work What this page rests on and what it can’t tell you · all 8 claims with their sources · the forensic record
Evidence Passport · public beta 8 addressable claims, 8 source dossiers, one permanent baseline. The passport is unreviewed — an AI-assisted structural migration, not a human or expert endorsement. Every revision keeps a permanent copy under revision history.
Inspect the claims, one by one — verbatim text, sources and caveats for all 8
The text below is copied verbatim from the published record. Each claim names the source dossiers that carry it; caveats stay attached to the claim they qualify.
clm-16c0647c96d61996
quantitative
Delhi's Public Works Department has installed roughly 280,000 street CCTV cameras since 2018, sited largely where resident welfare and market associations asked for them
Sources carrying this claim
clm-d04d02b31005567a
quantitative
in March 2025 the government ordered a physical audit after finding 263,000 recorded as installed, to check how many exist and still work
Sources carrying this claim
clm-f32d8aefd180d7c0
quantitative
Delhi Police runs a facial recognition system procured in 2018 under a High Court order to trace missing children, later used to identify people after the 2020 north-east Delhi riots and at protest sites
Sources carrying this claim
clm-5182c696affbf174
quantitative
Police treat an 80 percent similarity score as a positive match, and no Indian law specifically governs the technology
Sources carrying this claim
clm-90c63d505a95d315
quantitative
The Union Home Minister told Parliament in March 2020 that 1,922 faces had been identified with facial recognition after the north-east Delhi riots, matched against driving-licence and voter-ID data
Sources carrying this claim
clm-5b596dddf0a9a4c8
quantitative
In court filings Delhi Police had earlier put the system's accuracy at 2 per cent (2018) and under 1 per cent (2019)
Sources carrying this claim
clm-d4114efe9f3be2bf
quantitative
Delhi Police decided that an 80 per cent similarity score counts as a match, on a system whose accuracy their own court filings put at 2 per cent and then under 1 per cent
Sources carrying this claim
clm-67a6af698747b143
descriptive
which is why the government had to order a physical audit to find out how many of the cameras on its books exist and still work
Sources carrying this claim
Forensic record — source dossiers, gaps, hashes and the exact revision this page renders
This layer records what is known about each source, what is still missing, and the exact snapshot this page is rendering. A missing date, archive or hash stays visible as missing.
Awaiting operator and domain-expert review; publication of this revision implies neither.
Erratum on this revision's envelope. The snapshot, its hash and its permanent link are untouched.
- createdAt previously read 2026-08-27T13:00:00Z. The batching script that produced this revision advanced createdAt into the future to satisfy the strictly-increasing order the validator requires per entity. The value never recorded a real event; the corrected value is the moment the revision provably existed in the repository. Corrected to 2026-08-26T22:14:17Z on 2026-08-27; evidence: first appears in commit 788e6cc022, committed 2026-08-26T22:14:17Z.
sha256:d5baafb72708f05ec63db07b91d243abba5e9c7b319f31a345fd5e8547ce5913
- 8
- of 8 claims mapped
- 8
- source dossiers
- 7
- publication dates recorded
- 0
- archives linked
- 0
- exact locators
- 0
- content captures hashed
Governance facts carried by the record
- Operator
- Delhi Public Works Department (camera network); Delhi Police (facial recognition)
- Actor types
- city-government, national-government
- Funding model
- public
- Funding description
- Not recorded
- Cost
- Not recorded
- Ownership
- Not recorded
- Decision rights
- Not recorded
- Exit conditions
- Not recorded
8 source dossiers
PTI via The Week — Delhi PWD street CCTV programme (14 May 2025)
- Stable source ID
src-fe8b7a4c065f5774- Publisher
- Name not classified · theweek.in
- Type / language
- web-page · language not classified
- Published
- 2025-05-14
- Updated
- Not recorded
- Source last checked
- No source-level check date
- Record last verified
- 2026-08-27
- Dossier created
- 2026-08-27
- Retrieved copy
- No local capture recorded
- Archive
- No archived copy linked
- Content hash
- No captured content hash
- Locator
- Descriptive only — not page/paragraph exact
Open the recorded source note
14 May 2025 wire report: since 2018 the department has installed approximately 2.8 lakh CCTV cameras in two phases; ~2,000 cameras per assembly constituency; implemented with support from resident welfare and market trader associations.
- Claims linked
Business Standard — Delhi govt to conduct audit of over 260,000 CCTV cameras
- Stable source ID
src-f350b01b34ec14bf- Publisher
- Name not classified · business-standard.com
- Type / language
- web-page · language not classified
- Published
- 2025-03-31
- Updated
- Not recorded
- Source last checked
- No source-level check date
- Record last verified
- 2026-08-27
- Dossier created
- 2026-08-27
- Retrieved copy
- No local capture recorded
- Archive
- No archived copy linked
- Content hash
- No captured content hash
- Locator
- Descriptive only — not page/paragraph exact
Open the recorded source note
30 Mar 2025: 280,000-camera project, 263,000 shown installed in records; audit ordered to physically verify existence, function, image quality and data handling.
Comparitech — The World's Most Surveilled Cities (research)
- Stable source ID
src-4382828c9fc470a3- Publisher
- Name not classified · comparitech.com
- Type / language
- web-page · language not classified
- Published
- 2019-08-15
- Updated
- Not recorded
- Source last checked
- No source-level check date
- Record last verified
- 2026-08-27
- Dossier created
- 2026-08-27
- Retrieved copy
- No local capture recorded
- Archive
- No archived copy linked
- Content hash
- No captured content hash
- Locator
- Descriptive only — not page/paragraph exact
Open the recorded source note
Research ranking of camera density. Its Delhi figure is cited only as a comparison and is not used to support a claim in this entry; the publisher has since revised the underlying methodology.
- Claims linked
- Background source; no baseline claim points here
Outlook India — Delhi Police's 80 per cent match threshold raises concern, finds RTI
- Stable source ID
src-3a2f92f783812877- Publisher
- Name not classified · outlookindia.com
- Type / language
- web-page · language not classified
- Published
- 2022-08-18
- Updated
- Not recorded
- Source last checked
- No source-level check date
- Record last verified
- 2026-08-27
- Dossier created
- 2026-08-27
- Retrieved copy
- No local capture recorded
- Archive
- No archived copy linked
- Content hash
- No captured content hash
- Locator
- Descriptive only — not page/paragraph exact
Open the recorded source note
August 2022, on Internet Freedom Foundation RTI replies: a match is treated as positive at 80 per cent accuracy, below which it is a false positive; sub-threshold matches still pursued via corroborative evidence; use in the 2020 riots and later cases.
The Week — Facial recognition identified 1,900 Delhi rioters: how does the technology work
- Stable source ID
src-778d2d87933e26fc- Publisher
- Name not classified · theweek.in
- Type / language
- web-page · language not classified
- Published
- 2020-03-13
- Updated
- Not recorded
- Source last checked
- No source-level check date
- Record last verified
- 2026-08-27
- Dossier created
- 2026-08-27
- Retrieved copy
- No local capture recorded
- Archive
- No archived copy linked
- Content hash
- No captured content hash
- Locator
- Descriptive only — not page/paragraph exact
Open the recorded source note
13 Mar 2020: the Home Minister said 1,922 faces had been identified; driving-licence and voter-ID data used, no Aadhaar; carries the court-record accuracy figures of 2 per cent (2018) and under 1 per cent (2019).
- Claims linked
Panoptic (Internet Freedom Foundation project)
- Stable source ID
src-a81a67923fa33800- Publisher
- Name not classified · panoptic.in
- Type / language
- web-page · language not classified
- Published
- Not recorded
- Updated
- Not recorded
- Source last checked
- No source-level check date
- Record last verified
- 2026-08-27
- Dossier created
- 2026-08-27
- Retrieved copy
- No local capture recorded
- Archive
- No archived copy linked
- Content hash
- No captured content hash
- Locator
- Descriptive only — not page/paragraph exact
Open the recorded source note
Civil-society tracker of facial recognition deployments in India, including the Delhi Police system. Continuously maintained, so it states no publication date.
The News Minute — Explained: Delhi Police's AI surveillance network and why it's under scrutiny
- Stable source ID
src-5218213f0ef1853f- Publisher
- Name not classified · thenewsminute.com
- Type / language
- web-page · language not classified
- Published
- 2026-07-23
- Updated
- Not recorded
- Source last checked
- No source-level check date
- Record last verified
- 2026-08-27
- Dossier created
- 2026-08-27
- Retrieved copy
- No local capture recorded
- Archive
- No archived copy linked
- Content hash
- No captured content hash
- Locator
- Descriptive only — not page/paragraph exact
Open the recorded source note
2026 explainer: a Delhi Police order of 9 June 2022 as the sole authority for facial-recognition use; Safe City phase inaugurated February 2026 with 2,100 AI-enabled cameras live; notes India has no dedicated law and the broad state exemptions in the DPDP Act 2023.
- Claims linked
The Tribune — Facial recognition and AI surveillance vehicles deployed at Jantar Mantar protest site
- Stable source ID
src-38cd825c491c0468- Publisher
- Name not classified · tribuneindia.com
- Type / language
- web-page · language not classified
- Published
- 2026-07-24
- Updated
- Not recorded
- Source last checked
- No source-level check date
- Record last verified
- 2026-08-27
- Dossier created
- 2026-08-27
- Retrieved copy
- No local capture recorded
- Archive
- No archived copy linked
- Content hash
- No captured content hash
- Locator
- Descriptive only — not page/paragraph exact
Open the recorded source note
24 July 2026 currency check: facial recognition cameras and 360-degree surveillance vans deployed at a protest site; police describe it as routine practice also used on national holidays and at festivals.
- Claims linked
Known limits of this passport
- Publisher names and source languages have not yet been classified; domains are recorded without guessing.
- Existing source notes are descriptive locators, not page- or paragraph-exact citations.
- A missing archive, publication date or content hash is shown as null rather than inferred.
- The project-level evidence grade ranks the strongest source in the entry; it is not a truth score for every claim.
Continue exploring
Compare this project
Explore the wider context
📎 Cite this project
1001 Smart Cities (2026). “Citywide CCTV Rollout and Delhi Police Facial Recognition” — New Delhi, India. The Smart City Atlas. https://1001smartcities.org/projects/delhi-cctv-facial-recognition/ (last verified 2026-08-27). Data: CC BY 4.0.
The underlying data is free to reuse with attribution — see the open data page.
