Skip to main content
Version: 0.2 Draft 3

Namespace: x-ai-cn:

China AI Governance Regulations​

Jurisdiction: People's Republic of China (CN) Frameworks addressed: Provisions on the Administration of Algorithmic Recommendations (effective 1 March 2022); Provisions on the Administration of Deep Synthesis Internet Information Services (effective 10 January 2023); Interim Measures for the Management of Generative AI Services (effective 15 August 2023); Measures for Labeling AI-Generated and Synthetic Content (issued March 2025, effective 1 September 2025). Enforced by the Cyberspace Administration of China (CAC) jointly with MIIT and MPS. Registration tier: Open Review status: ai_reviewed Review date: 2026-03-20

Overview​

China regulates AI through separate instruments for algorithmic recommendations, deep synthesis, and generative AI services rather than a single overarching AI act. The 2025 Measures for Labeling AI-Generated and Synthetic Content (effective 1 September 2025) introduced a dedicated labelling regime. Algorithm filing requirements apply to certain recommendation services under the 2022 Provisions.

Fields​

FieldTypeRequiredDescription
x-ai-cn:generative_ai_service_flagbooleanOPTIONALSet to true where the AI system provides a generative AI service to the public within China under the 2023 Measures. Note: certain research and internal uses are outside scope under Article 2.
x-ai-cn:deep_synthesis_flagbooleanOPTIONALSet to true where the AI system uses deep synthesis technology to generate, edit, or synthesise content involving real individuals. Providers must obtain user consent, label generated content, and maintain content logs.
x-ai-cn:algorithm_recommendation_flagbooleanOPTIONALSet to true where the AI system uses algorithmic recommendation technology. Obligations include disclosure of recommendation logic, user choice rights, algorithm registration and filing for specified services, and labelling of algorithmically recommended content.
x-ai-cn:algorithm_filing_flagbooleanOPTIONALSet to true where algorithm filing or filing updates with the CAC are required or have been completed under the 2022 Algorithmic Recommendations Provisions.
x-ai-cn:security_assessment_typeenumOPTIONALThe type of security assessment applicable or completed. Permitted values: generative_ai_service, algorithm_recommendation, cross_border_data_transfer, none. Note: AI-service assessment and data export assessment are separate obligations under different instruments.
x-ai-cn:ai_content_label_measures_apply_flagbooleanOPTIONALSet to true where the 2025 Measures for Labeling AI-Generated and Synthetic Content (effective 1 September 2025) apply to content produced by this system.
x-ai-cn:content_labelling_flagbooleanOPTIONALSet to true where AI-generated or AI-assisted content produced by this system is labelled in accordance with applicable labelling requirements under the Deep Synthesis Provisions, Generative AI Measures, and the 2025 Labelling Measures.
x-ai-cn:training_data_compliance_flagbooleanOPTIONALSet to true where training data has been assessed for compliance with applicable Chinese law including consent, intellectual property, and content regulation requirements.

Revision notes (from ai_reviewed pass, 2026-03-20)​

  • ai_content_label_measures_apply_flag added referencing the 2025 Labelling Measures (effective 1 September 2025).
  • algorithm_filing_flag added reflecting filing requirements under the 2022 Provisions.
  • cac_security_assessment_completed_flag replaced by security_assessment_type to distinguish AI-service assessment from cross-border data transfer assessment.
  • content_labelling_flag description updated to reference all three relevant instruments including the 2025 Measures.
  • Algorithm recommendation description updated to include filing requirements and user choice rights.

Citations​

  • CAC Algorithmic Recommendation Provisions, effective 1 March 2022
  • CAC Deep Synthesis Provisions, effective 10 January 2023
  • CAC Interim Measures for Generative AI Services, effective 15 August 2023
  • CAC Measures for Labeling AI-Generated and Synthetic Content, March 2025, effective 1 September 2025

This namespace is registered and maintained by OPSF (Open Privacy Standards Foundation). Contact registry@pct.opsf.org for queries.