On the morning of September 6, at the forum of 2024Inclusion · The Bund Conference "Protecting AI with AI, the way to attack and defend in the era of big model", WDTA World Digital Technology Institute officially released the international standard "Security Requirements for Big Model Supply Chain". This standard is jointly prepared by experts from dozens of domestic and foreign units, such as Cloud Security Alliance Greater China, Ant Group, Microsoft, Google, Meta, PrivateAI, etc. The report was released by the Cloud Security Alliance in Greater China, and its member units China Telecom, Xi'an University of Electronic Science and Technology, Guangzhou University, Ant Group, Baidu and Tianyi Cloud jointly prepared it. It not only deeply analyzed the credibility of AI technology in practical applications, but also from the key dimensions of reliability, security, transparency and fairness, The advantages and challenges of current AI technology are comprehensively evaluated, and the strategy of AI system credibility is proposed, covering technical measures, policy recommendations and industry best practices.