Social risks and governance paths of generative artificial intelligence applications
The development and application of generative artificial intelligence will trigger a linkage transformation of social production relations,innovation methods,organizational structure,operational status,and value criteria.In the process of change,it can also lead to the infringement of user data privacy,the abuse of monopoly power by large technology companies,the generated content can be false,misleading,discriminatory,and may have other social risks with diffusion and penetration such as being difficult to determine the responsible party.Therefore,structured and systematic governance should be carried out for the design,application,supervision,and relief processes within the lifecycle of generative artificial intelligence.Firstly,it is necessary to establish an application list system,improve the management and circulation system of data grading and classification,and establish a data trainer access system to control the generation of related risks from the beginning.Secondly,improve the information content review and filtering mechanism,enhance the participation and intervention of regulatory agencies through technology embedding,and carry out full process supervision of generative artificial intelligence.Finally,a comprehensive damage relief system and a two-aspects responsibility allocation scheme should be established to minimize the negative impact of related damage consequences.
万方数据
