China plans to issue security regulatory standards covering the full AI lifecycle to prevent the risk of models becoming uncontrolled.

According to Bloomberg, Wang Ruomeng, an official at China’s National Development and Reform Commission, said on Tuesday that China is accelerating AI legislation and will implement full-lifecycle regulation of AI development and application, covering areas such as data and computing power. These standards aim to prevent the risk of increasingly powerful AI slipping out of control and are considered one of the most comprehensive AI safety guidelines available today. The regulatory framework will encompass major developers including DeepSeek, Moonshot, and Alibaba, which are working to create products capable of competing with the top-tier models from OpenAI and Anthropic.