Shifting the Dynamics: Analysis of the Impact of Generative AIs on the Use of E-Resources in Universities in Ghana

Abstract

Electronic resources (e-resources) have long been fundamental to academic research and scholarly communication, especially within university libraries. Recent advances in Generative Artificial Intelligence (Gen AI), accelerated by the COVID-19 pandemic and swift digital transformation, have begun to reshape information-seeking behaviours in higher education, particularly in resource-limited contexts. This study investigates how generative AI tools are affecting usage patterns of e-resources, the extent to which they may hinder or enhance access to scholarly materials, and users' perceptions of their integration within academic libraries in Ghana. Guided by the Technology Acceptance Model (TAM), the study explores perceived usefulness and perceived ease of use as key factors influencing AI adoption alongside traditional e-resources. A quantitative, cross-sectional survey design was used to collect data from 1,581 students across selected public, technical, and private universities in Ghana. A census approach ensured comprehensive representation of the target population. Data were gathered using a validated, self-developed questionnaire and analysed using descriptive and inferential statistics. The results show that although generative AI tools are steadily impacting academic information-seeking behaviour, they have not entirely replaced traditional e-resources. Although 58.0% of respondents indicated that they still rely on established databases, academic journals, and digital libraries, 42.0% reported a decline in their use of traditional electronic resources. This reduction was largely attributed to the perceived convenience, speed, and efficiency offered by generative AI tools. Further analysis revealed a modest but statistically significant effect (β = .18, p = .011), indicating that senior students are more likely to diversify their use of various information tools, whereas junior students mainly depend on ChatGPT and AI tools. The findings highlight the emergence of a hybrid information environment where generative AI acts more as a complementary facilitator rather than a substitute for authoritative scholarly resources.