Abstract This study examined human–AI collaboration in research assistance services in academic libraries in Cross River State, Nigeria. It investigated how researchers and students use AI tools, assessed librarians' perceptions of AI integration, identified adoption challenges, and proposed strategies, recommendations, and a framework for improving AI-supported research assistance. A mixed-methods design was adopted. The population comprised 161 academic librarians, postgraduate students, and faculty researchers from the University of Calabar, University of Cross River State, Cross River State Institute of Technology and Management, Ugep, and the College of Education, Akamkpa. Accidental sampling selected 120 respondents (80 librarians and 40 students/researchers). Data were collected through questionnaires and interview guides validated by experts in Library and Information Science and Test and Measurement; reliability yielded a Cronbach's Alpha of 0.76. Data were analyzed using frequencies, percentages, mean scores, and thematic analysis. Findings showed that ChatGPT (27.5%) and Grammarly (22.5%) were the most used AI tools. Librarians perceived AI positively, agreeing that it can enhance research assistance (Mean = 4.00) and should complement rather than replace librarians (Mean = 4.04). Poor ICT infrastructure (31.7%) and lack of training (26.7%) were major barriers to integration. It concluded that effective human–AI collaboration can improve research support services when supported by adequate infrastructure, AI literacy training, ethical guidelines, and institutional support. It recommended increased investment in digital infrastructure, continuous librarian AI training, and institutional policies for responsible AI use.
Keywords: Artificial Intelligence (AI), Research Assistance, Human–AI Collaboration
Publication Date: 2026-10-31