Ethical Acceptance Of Artificial Intelligence Surveillance

With the rapid development of big data and intelligent algorithms,artificial intelligence(AI)technology and its application have penetrated into all aspects of human society.AI surveillance technology has also gradually emerged,from "classroom monitoring" to”AI supervisor",which has aroused a heated discussion on the social risks and ethical issues of AI surveillance.On the one hand,AI surveillance brings benefits in terms of efficiency and security,on the other hand,it also brings some worrying problems such as citizen privacy,data security,algorithm discrimination.Therefore,the governance and supervision of AI surveillance is imminent.Traditional AI ethics and governance research is mainly carried out through expert interviews or questionnaire surveys,which is difficult to capture the real ideas of the public,research based on large-scale public comments mining is very rare.In this paper,we collected the blog,comments and Q&A data about AI surveillance in weibo and Zhihu through crawler,and divided them into education scene,work scene and life scene,finally obtained the research results through natural language processing technology and deep learning algorithm mining.The main research contents and achievements of this paper are as follows:1)Using syntactic dependency,EITS network sorting algorithm and word clustering algorithm to get the main concerns of the public about AI surveillance and the importance ranking of sub scenes,finally combined deontology and teleology to establish an ethical review framework for AI monitoring technology.2)Using PyTorch deep learning framework、LSTM long-term and short-term memory network and its derivative algorithm to calculate and analyze the sentiment of the comment text on each concern attribute,and get the attitude and sentiment of the public on each concern issue,which is of great significance to understand the main concerns and dissatisfaction of the public on AI surveillance,so as to better constrain technology and promote the benign development of technology.3)Using moral foundation theory(MFT)and Sentence Bert algorithm to calculate the public moral sensitivity and ethical acceptance,and its evolution over time,which is of great significance to understand the evolution of public moral concern.4)By using econometric regression model,this paper studies the influence of public attitude on ethical acceptance and the moderator mechanism of moral sensitivity.The results show that the education scene pays more attention to motivation,the work scene cares more about users and ways of use,and the life scene is more concerned about risks and benefits;the higher the moral sensitivity of a group,the greater impact of privacy and data issues,regulatory and governance issues,ways of use and users on their ethical acceptance.