高校图书馆应用生成式人工智能的风险与对策研究

Study on Risks and Countermeasures of Generative AI Application in University Libraries

摘要: 文章聚焦高校图书馆应用生成式人工智能的风险与对策，首先，阐述生成式人工智能赋能高校图书馆建设的场景，包括参考咨询服务升级、信息素养教育创新、科技查新效能提升、阅读推广服务优化及古籍信息处理革新。接着，分析应用过程中面临的风险，如AI幻觉风险、可版权性不明、权利归属争议、用户信息被不合理使用以及系统稳定运行风险。最后，提出对策：积极履行告知义务、协同发力缓解AI幻觉风险、关注跟踪立法动态和司法实践、做好权利义务分配、加强用户信息保护以及开展分类专项培训。

Abstract: This paper focuses on the risks and countermeasures of applying generative artificial intelligence (GAI) in university libraries. Firstly, it elaborates on the scenarios where GAI empowers the development of university libraries, including the upgrading of reference and consultation services, the innovation of information literacy education, the improvement of sci-tech novelty search efficiency, the optimization of reading promotion services, and the innovation of ancient book information processing. Secondly, it analyzes the risks encountered in the application process, such as the hallucination risk of generated content, unclear copyrightability, disputes over rights ownership, unreasonable use of user information, and the risk of unstable system operation. Finally, it puts forward countermeasures, including actively fulfilling the obligation of notification, making joint efforts to mitigate the risk of AI hallucination, paying attention to and tracking legislative developments and judicial practices, properly allocating rights and obligations, strengthening the protection of user information, and carrying out classified and specialized training.