As AI risks grow, China mulls how to make open-weight models less dangerous - South China Morning Post

China is working on ways to make openly available AI models safer and less prone to misuse. This matters because open-source AI tools are spreading globally, and how major countries like China approach safety could influence standards worldwide.