China Draft AI Law

Applies To

Overview

China has been developing a comprehensive AI law at the national legislative level through the National People's Congress Standing Committee (NPC-SC). The drafting process has proceeded in parallel with China's existing piecemeal AI regulatory instruments, including the Interim Measures for the Management of Generative AI Services, the Provisions on the Management of Algorithmic Recommendations, and the Measures for Labeling AI-Generated Synthetic Content. Unlike these sector-specific or technology-specific regulations, the proposed AI law is intended to provide a unified legal framework governing the full lifecycle of AI, from research and development through deployment and post-market monitoring. While no official draft text has been publicly released as of mid-2025, legislative planning documents and NPC-SC work plans have confirmed the law is in active development. Expected provisions include a risk classification system, obligations for AI providers and users, data governance requirements, liability allocation, and State supervision mechanisms. The law is anticipated to interact with and potentially supersede or consolidate aspects of existing AI-specific regulations. Enterprises operating in China or providing AI products and services to Chinese customers should treat the draft law as a material legislative risk requiring active monitoring. The timeline for finalisation and enactment remains uncertain.

Key Requirements

• •No binding obligations have been issued under this instrument as it has not yet been enacted
• •Expected to establish a tiered risk classification system for AI systems, analogous in structure to the EU AI Act
• •Anticipated to impose obligations on AI developers, providers, and deployers covering safety assessment, registration, documentation, and ongoing monitoring
• •Likely to include explicit liability provisions allocating responsibility among AI supply chain participants
• •Expected to require State approval or notification for high-risk AI systems prior to deployment
• •Anticipated data governance requirements will interact with existing obligations under the Personal Information Protection Law (PIPL) and Data Security Law (DSL)
• •Public sector and critical infrastructure AI applications expected to face heightened requirements

What Your Organization Must Do

• →Assign someone to monitor National People's Congress Standing Committee work plans for any published draft text.
• →Map your AI systems used in China, noting developer, provider and deployer roles for each.
• →Review current compliance with China's existing AI rules, since the law may consolidate them.
• →Check how your data governance already meets the Personal Information Protection Law and Data Security Law.
• →Treat the draft as a planning risk, not a current duty, in board and risk reporting.
• →Record contract positions on liability across AI suppliers, so allocation changes can be renegotiated quickly.

Playbook Guidance

Step-by-step implementation guidance for compliance teams.