China is preparing to roll out a broad set of standards governing the safety of artificial intelligence, from development through deployment. According to Bloomberg, they represent some of the most comprehensive guidelines yet proposed to guard against a loss of control over increasingly powerful systems.
A senior official with the National Development and Reform Commission, the country's state economic planner, said the government is working on legislation to govern AI safety at every stage. The planned framework would not be limited to the most advanced models, but would also cover the foundational elements that make such systems possible, such as training data.
The move follows a series of safety episodes worldwide that have fuelled the international debate on AI risks. In recent months more than 300 scientists from 32 countries have called for tighter controls, including air-gapped systems and a global body to log incidents, while Microsoft has proposed an emergency shutdown mechanism for the most powerful models.
China is no stranger to such measures: Beijing has already extended export controls to large models and their weights, with mandatory registration, as part of a broader strategy in its technology rivalry with Washington. If approved, the new safety rules would add to that regulatory framework.
It remains to be seen when the text will be finalised and what concrete obligations it will impose on companies. For now the broad picture suggests AI safety will be addressed not only at the level of final models, but across the entire chain of development and distribution.

No comments yet. Be the first!