Potential Safety Issues and Moral Hazard Posed by Artificial General Intelligence

Abstract. Artificial Intelligence (AI), a technology with a wide range of intelligence capabilities, has developed rapidly in recent years, bringing significant convenience and efficiency to society. However, most of the current artificial intelligence technologies belong to narrow artificial intelligence. Unlike Narrow AI, Artificial General Intelligence (AGI) possesses a more comprehensive understanding and problem-solving capability. AGI can learn in an unsupervised manner. General artificial intelligence can not only stand out in specific fields. It can also make effective decisions to a certain extent and operate in a wide range of environments. However, rapid progress has also raised widespread concerns about its potential risks. Therefore, the development of artificial intelligence requires standardization, which is urgent to ensure that it can make decisions that benefit humanity. Based on existing literature and data results, this paper explores the security issues and moral risks that general artificial intelligence may bring to humans. The research results indicate that these risks include user privacy breaches, system security issues, and social ethical conflicts. Dealing with these risks requires the joint efforts of all practitioners. This includes developing AGI in an ethical manner and ensuring that AI does not engage in activities that violate human interests.