Artificial Intelligence law at China
China has established a comprehensive regulatory framework for Artificial Intelligence (AI), emphasizing ideological alignment, content moderation, and sector-specific oversight. The Cyberspace Administration of China (CAC) serves as the primary regulatory authority, supported by various ministries overseeing specific sectors.(Home | White & Case LLP)
🧭 Key Regulatory Measures
1. Generative AI Measures (July 2023)
The CAC's Generative AI Measures set foundational rules for AI service providers:
Content Compliance: AI-generated content must adhere to laws and reflect "core socialist values," avoiding illegal, harmful, or misleading information.
User Registration: Mandatory real-name registration for users to prevent anonymous misuse.
Content Moderation: Providers must promptly address and report harmful content.
Penalties: Violations can result in fines, service suspension, or criminal charges .(Global Practice Guides, ChatonAI)
2. Draft AI Law (March 2024)
A scholarly draft proposes a comprehensive AI law covering:
Development and Promotion: Encouraging AI innovation.
Rights Protection: Safeguarding individual and organizational rights.
Security Obligations: Ensuring AI systems are secure and reliable.
Supervision and Management: Establishing regulatory oversight.
Special Application Scenarios: Addressing unique AI use cases.
International Cooperation: Promoting global AI governance .(Global Times)
3. AI Content Labeling and Watermarking
To combat misinformation, the CAC mandates:
Explicit Labels: Clear indicators on AI-generated content (text, audio, video, images).
Metadata Inclusion: Embedding identifiers in AI-generated files.
Platform Responsibility: Social media platforms must verify and label AI-generated posts, even without user declarations .(Global Practice Guides, MEDIANAMA, WIRED)
4. Sector-Specific Oversight
Different ministries oversee AI applications in various sectors:
Cyberspace Administration: General AI governance.
Ministry of Industry and Information Technology (MIIT): Standardization and industry development.
Ministry of Science and Technology: Ethical considerations and research oversight.
Public Security Bureau: Law enforcement and data security .(Reuters, Pillsbury Law)
⚖️ Legal Responsibilities and Liabilities
The draft AI law outlines liabilities for:
Developers and Providers: Accountable for AI system outputs and compliance.
Users: Responsible for lawful use of AI applications.
Penalties: Range from fines to criminal charges, depending on the severity of violations .(ICLG Business Reports, ChatonAI)
🌐 Ideological Alignment
AI models in China are required to:
Reflect Core Socialist Values: Align with Communist Party ideology.
Avoid Sensitive Topics: Exclude content related to politically sensitive issues.
Government Oversight: Models undergo government reviews to ensure compliance .(Financial Times, Reuters, Business Insider)
🧠 AI Model Development
China has developed several large language models, including:
Manus: An AI assistant featured on state media, developed in collaboration with Alibaba's Qwen AI models.
Xi Jinping Thought Model: A model trained on the ideology of President Xi Jinping .(Reuters, Time)
🔮 Future Outlook
China aims to finalize its AI law by 2025, aligning with its broader goal of becoming a global leader in AI technology. The regulatory framework balances innovation with stringent content control, reflecting the country's emphasis on ideological conformity and national security.
RELATED Blog
0 comments