Logical Updating and Institutional Optimisation of Artificial Intelligence Generated Content Labeling System
Generative artificial intelligence is disrupting people's access to information,but the quality of generated content is hard to distinguish.With the failure of the triage governance mechanism in the era of user-generated content,the labeling system should become the core of AI content governance,and play the role of dual-track prompting of"generation source+authenticity judgement".Model developers and service providers that occupy the information superiority position should fulfil the obligations of content safety guarantee and information quality disclosure,and take the initiative to correct the problems of user cognitive extrusion and psychological impact caused by false information.In the optimisation iteration of the specific marking system,the AI-generated content quality spectrum marking should be designed,requiring mandatory marking of the generation source,the responsible subject,and the content quality,in order to activate the substantive role of the marking system in the screening of information content.
generative artificial intelligencelabeling systemcontent governanceinformation identificationfalse information
万方数据
维普
