Generative AI-enabled Classified Protection for Library Reader Information
The development and application of generative AI provides new opportunities for library digitization,but it also brings challenges to the security of reader-related information,which manifests itself as the mishandling of reader information on a larger scale,with more complex behaviors and diverse scenarios.However,the existing reader information protection system has limitations,and there is an imbalance in the interests of reader information involved in generative AI.To solve these problems,it is necessary to develop classified protection rules according to the types of reader information defined in the Law on Public Libraries,so as to balance the claims of generative AI on patron information with the need to protect user information.Among them,stronger protection should be given to readers'private information,for which generative AI must strictly implement the rules of informed consent and purpose limitation;limited protection should be given to readers'personal information,the processing rules of which generative AI can flexibly follow;lesser protection should be given to library loan information,the processing of which by generative AI is encouraged,with obligations to prevent abuse of information protection.
generative AIreader informationbalance of interestsclassified protection
NETL
NSTL
万方数据
