Explainable AI and ethical information practices in library services

Abstract

This conceptual paper aims to examine the concept of explainable artificial intelligence (XAI) as a tool for maintaining information ethics in the context of library services. It investigates the importance of XAI in the context of artificial intelligence (AI) and responsible use of AI. A conceptual framework is developed based on the existing literature. Existing AI systems are examined to identify their limitations. To ensure the responsible use of AI, key concepts of XAI are analyzed, technological examples are examined, and a model for library services is proposed. This study argues that existing AI models have limitations and are not sufficiently transparent from the users’ point of view. Users may not know where the results come from or how the data have been analyzed. This uncertainty creates a lack of transparency in AI systems, which may undermine ethical practices. The study proposes that XAI systems can support the responsible use of AI and help ensure information ethics. This study offers a theoretical framework for understanding explainability and presents the key concepts related to the model. This framework can serve as a starting point for developing XAI-based library services. Libraries can use this model to design reference, book recommendation, and other relevant services.