China’s state economic planner announced on November 3 that it is accelerating work on comprehensive AI safety standards covering the full lifecycle of AI models, from development to deployment.
The standards will address data, algorithms and computing power. They respond to growing global concerns about AI security lapses and the potential loss of control over the technology.
The framework will apply to domestic AI developers, including Alibaba, DeepSeek and Moonshot AI. Developers will have to disclose safety risks and report incidents.
To ease the transition and support innovation, the government also plans to introduce a transition period and regulatory sandboxes where companies can test new technologies in controlled environments.