Three artificial-intelligence standards took effect simultaneously on November 1, forming a three-tier framework of evaluation, system and capability for AI governance.
They are GB/T 45654-2025 (basic security requirements for generative-AI services), GB/T 47863-2026 (compliance-management guide for AI service providers) and YD/T 7173-2026 (requirements for risk-management capability of AI security-governance systems).
The rollout tightens the compliance baseline for China's fast-expanding generative-AI and large-model industry, covering service security, provider obligations and organizational governance.
Source: China News Report (compiled from Toutiao AI Industry Hotspot).
