(Un)conventional ways of dialogic information retrieval using prompt engineering and the role of AI literacy

Abstract

Introduction. This study examines students’ use of ChatGPT for information retrieval and creative tasks, integrating AI literacy into library and information science (LIS) models and addressing gaps in AI-mediated processes and ethics. Method. A literature review and case study of 84 Jagiellonian University students, submitting 72 tasks, analysed transcripts, evaluations and authorship attributions. Analysis. Thematic analysis revealed eleven themes, including motivations, ethical issues, iterative strategies and AI literacy gaps. Strategies aligned with LIS models like berrypicking but exposed training deficiencies. Results. Most students (78%) claimed sole authorship; 18% cited co-authorship. Key strategies included iterative refinement (69%) and exploratory dialogues (67%). Outputs highlighted usefulness (82%), confidence (54%) and concerns over biases and authorship. Conclusions. Curricula must incorporate AI-specific competencies, such as prompt engineering, to promote ethical AI engagement. Future research should adapt frameworks like AUTOMAT and assess long-term AI literacy outcomes.

Keywords

• Artificial Intelligence in LIS, including ethical aspects
• Information literacy and related literacies
• Interactive information retrieval
• AI literacy
• dialogic information retrieval
• Generative Artificial Intelligence