China has released draft security standards for companies deploying generative AI services. The new regulations proposed by the National Information Security Standardization Committee may ban the use of certain types of data to train AI models, such as ones that “advocate terrorism” or “undermine national unity and social stability.”