Implementation Guidelines for the Standardised Application and Innovative Development of Intelligent Agents
Cyberspace Administration of China · 2026 · Originally Mandarin (WeChat)
China's national framework for safe and scalable agentic-AI deployment. Six sections, 38 numbered items: basic principles, technological foundations, security baseline (rule embedding, behavioural fencing, categorised and tiered governance), application-driven development across 19 sectors, innovation ecosystem, and implementation safeguards.
Only one translation so far. More external source documents will appear here as we engage with them.
Parallel work
MDSL has published a constructive-parallel framework as a v1.2 May 2026 draft: A Civil-Society Proposal for Sovereign and Federated Agentic AI in Aotearoa New Zealand →. It mirrors the structure of the CAC translation as a basis for comparison and dialogue.