V2I11P55

The Implementation and Impact of AI Chatbots in Academic Library Services

Herbert Wanga^1*

Abstract

This study explores how Artificial Intelligence (AI) chatbots are being used to improve academic library services in African universities. Using a systematic review of 42 peer-reviewed studies published between 2019 and 2024, together with regional policy reports, the paper examines both how chatbots are currently applied and the broader issues that shape their adoption. Five major chatbot platforms, ChatGPT, Gemini, Rasa, Claude, and Copilot, were compared using a context-aware evaluation model that considered cost, data privacy, language support, and infrastructure needs. The results show that while generative AI tools offer strong multilingual and conversational abilities, Rasa stands out as the most affordable, secure, and practical option for many African institutions facing bandwidth limitations and budget constraints. Building on these insights, the paper proposes a Hybrid AI Chatbot Framework that combines Rasa’s open-source NLU capabilities with secure generative AI extensions for handling more complex queries. This approach enhances data privacy, lowers operational costs, and supports bilingual English–Swahili interaction. It is also designed to align with key policies such as Tanzania’s Personal Data Protection Act (2022), the African Union Data Policy Framework (2022), and UNESCO’s AI Ethics Recommendation (2021). Although AI chatbots have the potential to expand access, improve responsiveness, and enhance user satisfaction, their adoption is still challenged by infrastructure gaps, policy inconsistencies, and ethical concerns such as bias and informed consent. The paper concludes with a practical roadmap and governance guidelines to support responsible, secure, and scalable chatbot deployment in African academic libraries.

Keywords:

Artificial Intelligence; Chatbots; Academic Libraries; Africa; Data Protection; Generative AI; Rasa; Hybrid Model; ICT Policy