关键词: 人工智能, 模型安全, 数据安全, 人工智能安全, 可信人工智能

Abstract: With the rapid popularization of automatic driving technology, intelligent assistant, face recognition,intelligent factory, smart city and other AI Artificial intelligence technologies and the rapid growth of related security incidents, consumers and the industry are paying more and more attention to AI network security issues and threats.The security of artificial intelligence applications and the dividends brought by intelligent applications, like two sides of a coin, always exist at the same time and the importance is highlighted.This paper discusses the development process of AI security in recent years and the security threats faced by AI application landing, and puts forward a set of reference security framework and landing implementation methods.

Key words: artificial intelligence, model security, data security, artificial intelligence security, trusted artificial intelligence