Artificial Intelligence Law of the People’s Republic of China (Draft for Suggestions from Scholars)

• Authority: Other authorities
• Collections: Chinese law and policy;Editors' Picks
• Most recent activity: Proposed
• Most recent activity date: 2024-03-16

Risk subdomain coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.

• Unfair discrimination and misrepresentation
• Exposure to toxic content
• Compromise of privacy
• AI system security vulnerabilities
• False or misleading information
• Disinformation, surveillance, and influence at scale
• Fraud, scams, and targeted manipulation
• Overreliance and unsafe use
• Loss of human agency and autonomy
• Environmental harm
• Lack of capability or robustness
• Lack of transparency or interpretability

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.

• Unequal performance across groups
• Pollution of information ecosystem
• Cyberattacks, weapon development or use
• Power centralization and unfair distribution
• Increased inequality and decline in employment
• Competitive dynamics
• AI pursuing its own goals
• AI possessing dangerous capabilities

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has good coverage of approximately 12-14 subdomains, with strong focus on discrimination and fairness (1.1, 1.3), privacy protection (2.1), AI system security (2.2), misinformation (3.1), malicious use prevention (4.1, 4.2, 4.3), overreliance (5.1), human agency (5.2), governance structures (6.5), and AI safety failures (7.1, 7.3, 7.4). Coverage is concentrated in discrimination/toxicity, privacy/security, malicious actors, human-computer interaction, and system safety domains.

Sector coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.

• Trade, Transportation & Utilities
• Information
• Finance & Insurance
• Scientific Research & Development Services
• Health Care & Social Assistance
• Public Administration (excl. National Security)
• National Security

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.

• Agriculture, Mining, Construction & Manufacturing
• Professional & Technical Services
• Management, Administrative & Support Services
• Educational Services
• Arts, Entertainment & Recreation

All other sectors had no coverage in this document.

Sectors Summary

This is a comprehensive national AI law that governs AI use across all economic sectors in China. It applies broadly to AI development, provision, and use activities across the entire economy, with specific provisions for government services, healthcare, finance, education, news/media, transportation (autonomous driving), and judicial applications. The law establishes cross-sectoral governance rather than sector-specific regulation.

Legislative Status

Hard Law

Legislative Status Summary

This is a draft legislative instrument with binding legal obligations, enforcement mechanisms including administrative penalties and fines, and mandatory compliance requirements indicated by extensive use of 'shall' language throughout.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Summary: The document is explicitly titled as a 'Draft for Suggestions from Scholars,' indicating it was proposed by academic experts. The document references state organs and the State Council as having authority to implement and coordinate AI governance.

Target

• Entity Types: AI_Developer, AI_Deployer, AI_User
• Summary: The document explicitly defines and regulates three categories of actors: AI developers (those who develop AI products and services), AI providers (those who provide AI products and services), and AI users (those who use AI products and services). These definitions are provided in Article 94 and obligations are specified throughout the document for each category.

Enforcer

• Entity Types: AI_Governance_Actor
• Summary: The document establishes 'main oversight departments for AI' as the primary enforcement bodies with authority to investigate violations, impose penalties, conduct inspections, and revoke licenses. People's courts handle litigation and people's procuratorates can file public interest lawsuits.

Monitor

• Entity Types: AI_Governance_Actor
• Summary: The document establishes monitoring responsibilities for main oversight departments for AI, including conducting risk assessments, random inspections, and security monitoring. Third-party organizations and expert committees provide assessment and monitoring support. Industry organizations conduct self-regulation monitoring.

AI Lifecycle Stages

Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems

• Plan and Design
• Collect and Process Data
• Build and Use Model
• Verify and Validate
• Deploy
• Operate and Monitor

AI Lifecycle Summary

The document comprehensively covers all AI lifecycle stages from planning through operational monitoring. It addresses design principles and planning (Articles 3-13), data collection and processing requirements (Articles 20-21, 45), model building and training (Articles 17-18, 51), verification and validation through risk assessments (Articles 42, 54), deployment requirements including registration (Article 53), and extensive operational monitoring obligations (Articles 41, 56, 63).

Technical Scope

Terms relating to technical scope that are mentioned or implicitly referenced in the document

• AI Models
• Foundation Models
• AI Systems
• Compute Threshold
• General Purpose AI
• Generative AI
• Open Weight / Open Source
• Predictive AI
• Task-specific AI

Technical Scope Summary

The document explicitly mentions AI models, AI systems, foundation models, and general purpose artificial intelligence (AGI). It establishes compute thresholds for critical AI classification based on parameters and scale. The document does not explicitly distinguish between generative AI and predictive AI, nor does it specifically address open-weight/open-source models or general purpose AI (GPAI) as distinct categories, though it does address open-source ecosystems generally.