Guest editorial: Application of artificial intelligence in information services: part 2

Application of artificial intelligence in information services: part 2Building upon the foundations laid in Part One (Issue 43.5), which explored the initial promises and entry points of artificial intelligence (AI), this collection presents a more complex and mature phase of the discourse.The ten papers featured here move beyond asking if or what AI can do, to critically examine how it is being integrated, for whom and with what consequences.They collectively paint a picture of a field in transition, grappling with the practical, pedagogical, ethical and epistemological realities of weaving AI into the very fabric of library and information work.This special issue reflects a significant broadening of scope.The studies encompass global educational reforms, evaluations of historical knowledge, user behavior modeling, public library case studies and novel applications in classification and robotics.A unifying thread, however, is the shift from viewing AI as a set of external tools to understanding it as a transformative force that reconfigures relationships: between teachers and students, between historical and contemporary knowledge, between users and systems and between libraries and their communities. Reconfiguring foundations: education and epistemologyThe transformation begins at the source of professional identity: education.The paper "Reconfiguring global LIS education: AI integration, pedagogical roles, and epistemic challenges" provides a sobering cross-continental analysis.It reveals that the integration of AI into library and information science (LIS) curricula is uneven, often creating a disconnect between technical skill acquisition and critical, ethical reflection.The authors warn of a worrying trend towards "role alienation," where educators risk becoming mere "technical facilitators" and students "system operators."This powerful analysis issues a clarion call: LIS education must evolve to cultivate not just tool operators, but "information governors" capable of auditing, critiquing and shaping the algorithmic systems that will define our future information landscape.Complementing this macro-view of knowledge dissemination is a micro-analysis of knowledge itself."Automated learning support literature classification using large language models via different strategies" reimagines the ancient library science task of classification through a cognitive lens.By developing a learning support literature classification (LSLC) scheme and testing various large language models (LLMs) for automation, the study does more than demonstrate technical feasibility (which it does successfully).It proposes a fundamental re-orientation: classification systems should be designed not just for inventory, but to actively scaffold human learning during search processes.This bridges core information science theory with cutting-edge AI, pointing toward more cognitively resonant knowledge organization. Evaluating impact and value in new dimensionsA critical function of our field is assessment, and AI is enabling novel frameworks for assessment."Contemporary technological value of ancient Chinese medical books: An empirical patent citation analysis using large language models" showcases this brilliantly.