KORU CHAPTER 02 · PUBLIC-SPACE EVIDENCE
This is the public-space chapter of KORU: a source-led record of what is documented after a person is seen, how that record may travel, and what is meant to constrain it.
ENTER THE PUBLIC RECORD
This chapter is built for different kinds of public scrutiny. Begin with the human question, the place, or the evidence you can help improve. Then follow the source trail at your own depth.
See how a public moment can become a connected, analysed, and governed record, and why a camera count alone cannot answer that question.
Read the chapterSearch the pilot, open a record, compare its layers, and trace every claim back to a dated public source.
Open the monitorSubmit it privately for editorial review. Only accepted, scoped changes become part of the public evidence history.
Submit evidenceFollow the record, not a headline
The monitor’s visual journey moves from the place people inhabit to the systems, rules, and sources that make public scrutiny possible.
THE PUBLIC-SPACE RECORD
Hover or select a record to move from a comparative view to the underlying measurement, source year, and confidence note.
The further right, the greater the camera intensity reported for a specific reference place. The higher, the stronger the measured protection from arbitrary privacy interference.
COUNTRY EVIDENCE LEDGER
| Rank | Country record | Reference evidence | Cameras / 1k | Privacy rights | Concern signal | Record |
|---|---|---|---|---|---|---|
| 01 | ChinaNational estimate | National estimate700,000,000 reported cameras | 494.25reference-place only | 9.7+0.0 pts | High concern | |
| 02 | IndiaReference-city estimate | Hyderabad900,000 reported cameras | 79.38reference-place only | 38.7-0.5 pts | Elevated concern | |
| 03 | PakistanReference-city estimate | Lahore410,297 reported cameras | 27.67reference-place only | 15.8-4.2 pts | High concern | |
| 04 | South KoreaReference-city estimate | Seoul243,426 reported cameras | 24.28reference-place only | 82.9-1.0 pts | Watch | |
| 05 | RussiaReference-city estimate | Moscow250,000 reported cameras | 19.63reference-place only | 8.9-3.3 pts | High concern | |
| 06 | AfghanistanReference-city estimate | Kabul90,000 reported cameras | 18.45reference-place only | 11.9-2.5 pts | High concern | |
| 07 | SingaporeCity-state estimate | Singapore112,999 reported cameras | 18.35reference-place only | 57.1+0.4 pts | Watch | |
| 08 | United KingdomReference-city estimate | London130,000 reported cameras | 13.40reference-place only | 82.8-0.2 pts | Watch | |
| 09 | United StatesReference-city estimate | Los Angeles46,766 reported cameras | 12.40reference-place only | 72.1-7.0 pts | Watch | |
| 10 | MexicoReference-city estimate | Mexico City136,920 reported cameras | 6.02reference-place only | 46.6-1.9 pts | Watch | |
| 11 | BrazilReference-city estimate | São Paulo51,346 reported cameras | 2.23reference-place only | 59.0+2.4 pts | Context needed | |
| 12 | EgyptReference-city estimate | Cairo48,033 reported cameras | 2.08reference-place only | 9.8-0.3 pts | Watch | |
| 13 | JapanReference-city estimate | Tokyo49,841 reported cameras | 1.35reference-place only | 86.6-3.2 pts | Context needed | |
| 14 | BangladeshReference-city estimate | Dhaka3,512 reported cameras | 0.14reference-place only | 7.9+1.8 pts | Watch |
FIVE SUPPLEMENTAL LAYERS
These layers explain the capability, connection, and conditions around a documented record without creating a second opaque score. A gap in the ledger means public evidence was insufficient for this pilot, not that the practice is absent.
Operational or live use supported by a public source in a named context.
Application, setting, decision role, scale, and stated safeguards kept visibly separate.
Network, local integration, or owner-mediated footage sharing; mechanisms remain distinct.
Hosting context, residency, public cloud, and lawful-access conditions, kept separate from surveillance evidence.
Named supplier, public body, function, and source; all other country files retain a visible evidence gap.
EVIDENCE LEDGER
Evidence of public-authority facial or biometric identification in a named public-space or policing context. This is a documented-presence layer, not an intensity score.
| Country record | Status | Named context | Evidence date | Confidence | Open source | Detail |
|---|---|---|---|---|---|---|
| ChinaCCTV reference: National estimate | Documented active use | National programme evidence | 2021 | Medium | CSET · 2021 | |
| IndiaCCTV reference: Hyderabad | Documented active use | Hyderabad, Telangana | 2021 | High | Amnesty International · 2021 | |
| PakistanCCTV reference: Lahore | Documented active use | Safe City networks in major cities | 2016–2025 | Medium | Biometric Update · 2016 | |
| South KoreaCCTV reference: Seoul | Documented pilot or limited use | Seoul public-health pilot | 2021 | Medium | Reuters · 2021 | |
| RussiaCCTV reference: Moscow | Documented active use | Moscow | 2021–2023 | Medium | Human Rights Watch · 2021 | |
| AfghanistanCCTV reference: Kabul | Documented active use | Kabul | 2025-02 | High | BBC · 2025 | |
| SingaporeCCTV reference: Singapore | Documented pilot or limited use | Smart Nation sensor programme | 2018 | Medium | Reuters · 2018 | |
| United KingdomCCTV reference: London | Documented active use | 13 police forces using or having used LFR | 2025-11 | High | Home Office · 2025 | |
| United StatesCCTV reference: Los Angeles | Documented active use | Federal law-enforcement agencies | 2024-03 | High | U.S. GAO · 2024 | |
| MexicoCCTV reference: Mexico City | Documented active use | Mexico City and state-level deployments | 2025 | Medium | R3D · 2025 | |
| BrazilCCTV reference: São Paulo | Documented active use | São Paulo Smart Sampa | 2026 | High | Prefeitura de São Paulo | |
| EgyptCCTV reference: Cairo | Documented active use | New Administrative Capital | 2023 | Medium | Reuters · 2023 | |
| JapanCCTV reference: Tokyo | Documented pilot or limited use | Police / event-security pilots | 2024 | Low | Personal Information Protection Commission | |
| BangladeshCCTV reference: Dhaka | No comparable public evidence located | Pilot evidence review | 2026-08 | Low | Evidence gap retained |
EVIDENCE-GAP MAP
This map shows the **completeness of the public record** across the pilot’s CCTV, AI, facial-recognition, and camera-integration layers. It does not visualise surveillance intensity or risk.
METHOD, VERSION 1.6
The pilot refuses a universal “most surveilled country” score because globally consistent public-camera counts do not exist. It pairs documented CCTV evidence with privacy protection in practice, then adds AI scope, facial recognition, private-camera links, data infrastructure, and supplier/procurement records as separate, source-specific context layers.
Capture the named city or country estimate, data year, public/private scope, and source limitation.
Pair it with WJP Factor 4.6, which measures protection from arbitrary interference with privacy.
Record the use case, setting, decision role, stated scale, and safeguards without translating them into a single AI score.
Record facial recognition and distinguish a voluntary footage request from a networked camera ecosystem, while preserving gaps.
Show hosting context, residency, public-cloud governance, and lawful-access rules without turning any of them into proof of access.
Record a supplier only when a public source also names the system or function and public body; retain the gap when it does not.
HOW THIS MONITOR CHANGES
Most camera and safeguards data changes in annual or irregular releases. The monitor should display those source vintages, record deltas, and tell visitors exactly when a figure was checked.
Keep the prior value, source date, and editorial note. The WJP privacy badge already shows the 2024–25 point movement in this pilot.
Audit direct CCTV sources, city boundaries, private-network integration, and unverified claims. Add records only when the underlying source can be opened.
Replace the WJP vintage when its annual release arrives. Add V-Dem and Freedom House as separate context layers, never as hidden proxies for camera counts.
SOURCE LEDGER
This first release uses only sources that can be opened by the public. Each source performs a different role, and none alone is a complete surveillance measurement.
Reference-city camera estimates, public-versus-private scope notes, and data-date caveats.
A documented-presence snapshot of government AI surveillance, including facial recognition. Used for corroboration, not a score.
Official, municipal, regulator, rights-group, and credible reporting sources for named facial-recognition contexts.
Country-level public sources that distinguish application, operating setting, decision role, scale, and documented human controls or gaps.
Public records that distinguish footage requests, camera registration, and structured integration from blanket system access.
Named supplier, public body, surveillance-relevant system, source type, and scope notes. A record is retained only where this chain is documented.
Bounded public records of surveillance-related incidents, formal challenges, findings, remedies, reforms, and evidence gaps.
Primary or qualified public records on data-centre policy and capacity. Capacity provides context only; it does not show who controls data.
Public-sector cloud frameworks and data-residency conditions, shown independently from surveillance records and facility capacity.
Statutory and procedural disclosure powers that may shape provider obligations; legal authority is never presented as evidence of use.
Factor 4.6: freedom from arbitrary interference with privacy, scored in practice.
Annual, expert-coded country-year context on government monitoring and privacy-law conditions.
The monitor’s rights test: legality, necessity, proportionality, and protection against arbitrary interference.
