Record STMT-WAB4709Unknown

Interim Measures for the Management of Generative Artificial Intelligence Services

生成式人工智能服务管理暂行办法

Regulation UnknownEastern Asia · CNin force

The Interim Measures for the Management of Generative Artificial Intelligence Services regulate generative AI services provided to the public within the People's Republic of China. The regulation emphasizes adherence to core socialist values, outlines provider responsibilities regarding data quality and security, and establishes safety assessment and content moderation mechanisms. It also encourages domestic innovation in AI algorithms, frameworks, and chips while penalizing illegal generation and dissemination of content.

Summary and content fields are drawn from the document text and are pending editorial review.

Who issued it

Cybersecurity Administration of ChinaRegulator · CN

Lead issuer

National Development and Reform CommissionGovernment (executive) · CN

Co-issuer

Ministry of Education of the People's Republic of ChinaGovernment (executive) · CN

Co-issuer

Ministry of Science and Technology of the People's Republic of ChinaGovernment (executive) · CN

Co-issuer

Ministry of Industry and Information TechnologyGovernment (executive) · CN

Co-issuer

Ministry of Public SecurityGovernment (executive) · CN

Co-issuer

State Administration of Radio and TelevisionRegulator · CN

Co-issuer

Legal forceUnknown

Under whose authorityUnknown

Judged fromfull text

Who it governs

General publicCompaniesAI developers and providersAI deployers and users

AI service providers, developers, and enterprises deploying generative artificial intelligence systems for public use in China must read this to ensure regulatory compliance.

Sectors

Cross-sectorTelecommunications and digital services · generative artificial intelligence services

What the document contains

What's covered

The document details requirements for data labeling, model training, safety assessments, user authentication, and reporting mechanisms for generative AI services. It lays out supportive policies for infrastructure development alongside punitive measures for security and legal violations.

What's missing

Technical specifications for testing model robustness or detailed thresholds for computational power exemptions are not stated.

Key takeaways

Generative AI services provided to the public within China must adhere to core socialist values and laws.
Providers are legally responsible for the legitimacy of pre-training data and optimization data used in generative AI products.
The regulation encourages independent innovation in basic technologies such as chips, software, algorithms, and platforms.
Providers must implement effective safety measures, including labeling generated content and handling illegal outputs promptly.
Supervisory authorities across multiple government departments hold enforcement powers regarding compliance and security assessments.

Principles it states

Adherence to core socialist valuesRespect for intellectual property rights and business ethicsProtection of personal information and prevention of illegal profilingEncouragement of safe, inclusive, and equitable AI development

Outline

· Chapter I: General Provisions
· Chapter II: Encouragement and Support Measures
· Chapter III: Basic Norms for Service Provision
· Chapter IV: Supervision, Inspection and Legal Liability
· Chapter V: Supplementary Provisions

Issued13 July 2023

Length206 words

FormatNot recorded

generative_ai content_moderation data_security algorithm_governance intellectual_property