Deconstructing the CCPLet the world understand the CCP. The CCP does not represent the Chinese people.

Institution

Police Big-Data Fusion Platforms: How Information Enters Policing

Cross-database aggregation, risk tags, lead dispatch, grassroots checks, and feedback loops.

Digital Governance, Censorship, and SurveillanceHuman Rights, Ethnicity, Religion, and RepressionData surveillanceSecuritizationVisibility controlMinistry of Public Security and the National Police Data SystemStreet-Level Stability Centers, Communities, and Grid SystemsNationwide Surveillance Infrastructure and Offline Enforcement
Contents

Visual Guide

Operational chain: Police Big-Data Fusion Platforms: How Information Enters Policing

Read from information intake to organizational consequence.

Stage 1Administrative registration, commercial platforms, and police systems continuously supply structured data.
→
Stage 2Identity numbers, devices, phone numbers, addresses, and relationship graphs link databases to a person.
→
Stage 3Rules or models generate anomaly, key-person, and associated-risk labels.
→
Stage 4Command platforms dispatch leads to police stations, communities, or special teams.

Information entering a police decision

A national identification number sits in household-registration records. A phone number appears in communications and platform accounts. An address enters community files, while vehicle, lodging, travel, and case records remain in other databases. A police fusion platform brings those records into a shared analytical environment, decides whether they refer to one person, and assembles relationships and past activity into an operational file. Once linked, ordinary movements can be reclassified as anomalies or associated risk.

The Personal Information Protection Law requires state bodies to process personal data within statutory authority, procedure, and necessity. The Data Security Law requires police data requests to follow national rules and approval procedures [3] [4]. These are formal limits. The public rarely sees field inventories, query logs, model rules, or internal authorizations, making it difficult to test whether a particular link was necessary, accurate, or lawful.

Aggregation and tags

Fusion begins at the data layer. Household registration, public-order, criminal-investigation, and immigration records can be connected through stable identifiers. External inputs may come from other government bodies, public video systems, and companies. The 2025 regulation on public-security video-image systems governs their construction and use and assigns supervisory duties to police [1]. Video is one input, but the regulation shows how images from social space become callable police resources.

Rules, lists, or models then assign labels. A label may describe an identity category, earlier police contact, a shared journey, or an anomaly requiring a check. It is a classification produced from selected fields, not an adjudicated fact. In an opaque police system, however, a local officer often sees the label before seeking an explanation. A stale address, incorrect entry, or relationship by association can become a new lead.

Dispatch and feedback

The mobile application for Xinjiang's Integrated Joint Operations Platform provides unusual evidence from near the operational end. Human Rights Watch's reverse engineering identified interfaces for person searches, individual files, information collection, and the transmission of tasks or leads [7]. The researchers did not log into the back-end servers and could not reconstruct every database or algorithm. The client still showed a workflow connecting collection, classification, and checks on people.

After a lead reaches a police station or another front-line unit, personnel can contact the subject, verify an address, ask about activity, and add information. The result returns to the system, where it may alter an old tag or add contacts and conduct. This feedback loop makes the platform more than an archive. Each offline encounter expands the system, and people around the subject may enter later scrutiny through association.

Xinjiang and platforms elsewhere

The UN human rights office's 2022 assessment found that available information indicated extensive aggregation by IJOP and the capacity to flag people as potential detention candidates [10]. That finding concerns Xinjiang's coercive environment. It does not establish identical fields, algorithms, or enforcement intensity nationwide. Procurement, acceptance records, and case files are needed for other localities.

The CCP has nevertheless expanded real-name account rules, algorithm filings, video governance, and police data capacity across China [5] [6]. The direction is clear: more daily records become identifiable, linkable, and deliverable to governance systems. Local variation does not remove the power imbalance. Police define data needs and possess coercive authority; companies and neighborhood bodies add entry points, while flagged people lack equal access to reasons and correction.

Disputes and evidence limits

The record establishes capabilities for cross-database linkage, personal files, task transmission, and information feedback. It also documents the relationship between data-driven policing and detention risk in Xinjiang. It is reasonable to conclude that tag quality and relational inference affect offline checks. The public evidence does not show that every locality uses one model, that all platform data is continuously open to police, or that every alert produces punishment. By withholding interfaces, false-positive rates, and appeal outcomes, the CCP makes citizens bear the uncertainty created by its systems.

Sources

  • [3] Personal Information Protection Law of the PRC.
  • [4] Data Security Law of the PRC.
  • [1] Regulation on public-security video-image information systems.
  • [7] Human Rights Watch reverse engineering of the IJOP app.
  • [10] UN human rights office assessment of Xinjiang.
  • [5] Rules on internet user account information.
  • [6] Rules on algorithmic recommendation services.

Key evidence

What the available sources establish

Official finding

The OHCHR assessment concluded that large-scale arbitrary detention and related abuses in Xinjiang may constitute international crimes, while individual responsibility requires further independent investigation.

Sources

  1. Regulation on Public Security Video Image Information Systemsprimary-recordLive
  2. MPS Rules for Public Security Video Information Systemsprimary-recordUnavailable
  3. Personal Information Protection Law of the PRCprimary-recordLive
  4. Data Security Law of the PRCprimary-recordLive
  5. Provisions on the Administration of Internet User Account Informationprimary-recordLive
  6. Provisions on Algorithmic Recommendation in Internet Information Servicesprimary-recordLive
  7. China's Algorithms of Repression: Reverse Engineering a Xinjiang Police Apptechnical-researchLive
  8. We Chat, They Watchtechnical-researchLive
  9. Censored Contagion IItechnical-researchLive
  10. OHCHR Assessment of Human Rights Concerns in Xinjiangofficial-findingLive
  11. Treasury Sanctions on Biometric Surveillance Technologyofficial-findingLive
  12. 2024 Country Report on Human Rights Practices: Chinagovernment-reportLive
  13. Official Accountability Record on the Henan Red-Code Incidentprimary-recordRedirected
  14. Investigation into Red Health Codes Assigned to Henan Bank Depositorsinvestigative-reportingLive
  15. CECC 2025 Annual Reportgovernment-reportLive
  16. Human Rights Watch Report on Detained White Paper Protestersinvestigative-reportingLive
  17. Amnesty International Interviews One Year after the White Paper Movementinvestigative-reportingLive

Related Reading

AnalysisThe Big-Picture Template: Sacrificing Rights To An Abstract CollectiveInstitutionData Governance And Social Control: Technology In Stability MaintenanceMechanismThe Interface Between Propaganda And Stability MaintenanceOverviewHow Technical Censorship Converges With Propaganda, Repression, And Rights Abuses