Regulates algorithmic recommendation services in China, emphasizing security, transparency, and rights of users. Mandates providers prevent harmful information dissemination, protect user rights, and ensure algorithmic accountability, with penalties for violations.
Analysis summaries, actor details, and coverage mappings were LLM-classified and may contain errors.
This is a binding legal regulation issued by Chinese government authorities with explicit enforcement mechanisms, mandatory compliance requirements, administrative penalties, and formal filing procedures.
The document has good coverage of approximately 12-14 subdomains, with strong focus on malicious actors (4.1, 4.2, 4.3), discrimination and toxicity (1.1, 1.2), privacy and security (2.1, 2.2), misinformation (3.1, 3.2), human-computer interaction (5.1, 5.2), socioeconomic issues (6.1, 6.2, 6.4), and AI system safety (7.1, 7.3, 7.4). Coverage is concentrated in content safety, user protection, algorithmic accountability, and preventing malicious use.
The document primarily governs the Information sector, specifically Internet information services, algorithmic recommendation platforms, and digital content providers. It also has significant coverage of Professional and Technical Services (through consultation and technical support requirements) and extends to multiple other sectors where algorithmic recommendation services are deployed, including Finance and Insurance, Health Care, Educational Services, Arts and Entertainment, Accommodation and Food Services, and work dispatch/gig economy platforms.
The document primarily focuses on the Deploy and Operate and Monitor stages of the AI lifecycle, with significant coverage of verification and validation requirements. It addresses algorithmic recommendation services that are already developed and emphasizes ongoing operational requirements, transparency, security assessments, and monitoring obligations.
The document explicitly covers algorithmic recommendation services and AI systems but does not specifically mention AI models as a distinct concept. It addresses algorithmic recommendation technology including generative/synthetic types, personalized recommendation, ranking, search filtering, and dispatching types. There is no mention of frontier AI, general purpose AI, foundation models, or compute thresholds. The document does address synthetic/generated content requiring labeling.
Cyberspace Administration of China, Ministry of Industry and Information Technology, Ministry of Public Security, State Administration of Market Regulation
The document is formulated and interpreted by multiple Chinese government agencies as stated in Articles 34 and 35, with the national cybersecurity and informatization department (Cyberspace Administration) responsible for overall coordination.
Cyberspace Administration of China, Ministry of Industry and Information Technology, Ministry of Public Security, State Administration of Market Regulation, local cybersecurity and informatization departments
Multiple government agencies are designated with enforcement authority including issuing warnings, fines, ordering rectification, and conducting security assessments as detailed in Articles 3, 28, 31, and 32.
Cyberspace Administration of China, telecommunications departments, public security departments, market regulation departments, local cybersecurity and informatization departments
The same government agencies responsible for enforcement also conduct ongoing monitoring through security assessments, supervision, inspections, and complaint handling as specified in Articles 23, 28, and 30.
Algorithmic recommendation service providers, Internet information service providers
The document explicitly targets 'algorithmic recommendation service providers' who use algorithmic recommendation technology to provide Internet information services within mainland China, as defined in Article 2.
18 subdomains (14 Good, 4 Minimal)