“Artificial Intelligence in Digital Libraries: Enhancing User Experience, Information Retrieval, and Service Delivery”

Abstract

Artificial intelligence (AI) is rapidly changing the architecture of digital libraries by shifting user interaction from conventional keyword-based retrieval toward conversational search, recommendation, automated metadata enrichment, semantic discovery, and adaptive service delivery. This paper examines how these changes affect user experience (UX), with particular attention to usefulness, usability, trust, accessibility, personalization, information quality, privacy, and human oversight. A structured evidence review was conducted using 30 authenticated scholarly and institutional sources spanning foundational digital-library evaluation research and contemporary AI-in-library studies. Evidence was synthesized from empirical surveys, mixed-method studies, systematic and scoping reviews, user-evaluation research, human-AI interaction scholarship, and international governance frameworks. Results indicate that AI can improve discovery, responsiveness, personalization, and service availability, but these gains are conditional rather than automatic. Published evidence shows strong interest in AI-enabled services, alongside persistent concerns about insufficient training, opaque algorithms, uneven collection coverage, privacy disclosure, bias, and the reliability of generated answers. In an environmental scan of 31 academic-library chatbots, 94% were AI-powered or hybrid, 42% supported multiple languages, but only 13% disclosed privacy policies. A North American survey of 242 librarians found that 67% expected AI-related technologies to transform library functions and 68% expressed interest in training. A recent systematic review found that funding was cited as a major institutional challenge in 48.28% of the 29 included studies. Across the evidence base, the most consistent UX determinants were relevance, ease of use, system functionality, transparency, information quality, and the availability of human escalation.