China's comprehensive AI Law expected to advance to second reading at the NPC during 2026.
China's national AI Law is expected to complete its second NPC reading in 2026, moving the country from administrative-regulation fragmentation (the 2022 Algorithm Recommendation Provisions, 2023 Deep Synthesis Provisions, 2023 Generative AI Interim Measures) to a unified statutory framework. The draft law, as published for consultation, introduces a tiered risk classification (general, significant, and critical AI) modelled partly on the EU AI Act's risk-based approach but with distinctly Chinese characteristics: stronger mandatory state-review provisions, enhanced data-localisation requirements, and an explicit national-security carve-out.
For international companies operating AI systems in mainland China — or integrating with Chinese AI service providers — the law's passage will trigger immediate compliance obligations including: registration of significant-risk AI systems with the CAC or MIIT, technical documentation requirements analogous to (but distinct from) the EU's Article 11, and mandatory human-oversight provisions for critical-infrastructure AI.
AIMenta tracks the China AI Law's legislative progress and has retained specialist PRC counsel to advise clients on the interaction between the new law and the sector-specific AI guidance already issued by PBOC (financial services), NHSA (healthcare), and SAMR (market regulation). Enterprises operating in both the EU and PRC will face the world's most complex dual-regulatory AI environment by 2027, and planning should begin now.