China releases version 2.0 of AI security governance framework
ChatGPT said:
China has released Version 2.0 of its AI Security Governance Framework, an updated set of rules to manage the risks of artificial intelligence as the technology evolves. Building on the 2024 framework, the new version refines risk classifications, introduces graded levels of risk, and updates preventive measures.
On 15 September 2025, during the National Cybersecurity Awareness Week, China’s Cyberspace Administration unveiled Version 2.0 of its Framework for Security Governance of Artificial Intelligence. The updated framework is designed to strengthen oversight of artificial intelligence (AI) as the technology develops rapidly and poses new risks.
The first version of the framework, released in September 2024, was part of China’s efforts to implement its Global Initiative on AI Governance. Over the past year, AI applications have expanded significantly, prompting authorities to refine their approach. Version 2.0 was jointly developed by the National Internet Emergency Response Centre, together with AI professional organisations, research institutes, and industry players.
According to Chinese officials, the new framework builds on the original by introducing refined risk classifications, exploring a system of graded risks, and setting out updated prevention and governance measures. The document is intended to track evolving risks in real time, ensuring that governance adapts to the fast pace of AI innovation.
Authorities say the framework reflects international trends in AI governance, aiming to balance technological progress with security, ethics, and trust. It emphasises building a development ecosystem that is ‘secure, trustworthy, and controllable,’ while also fostering international cooperation. Officials added that the framework will support multilateral efforts to share AI benefits globally, ensuring that advances in the field contribute positively to societies worldwide.
In practice, Version 2.0 signals China’s intention to link AI innovation more closely with oversight, setting standards for industry while positioning itself as a global voice in shaping AI governance.
