Powering Reference Transformation: A Global Analysis of Chatbot Adoption and Ethics in Academic Libraries

Abstract

As academic libraries adapt to evolving user expectations and increasingly digital service environments, chatbots are emerging as tools to enhance reference and frontline support services. This poster presents findings from a global environmental scan and literature review examining chatbot adoption in academic libraries. Thirty-one library chatbots were analyzed across diverse geographic regions, institutional contexts, and levels of technological maturity, including rule-based, AI-powered, and hybrid systems.

Using a structured set of service, information-seeking, and conversational queries, the study evaluated chatbot responsiveness, semantic understanding, service scope, and ethical transparency. Findings indicate that adoption remains limited and uneven globally. While most chatbots responded effectively to basic service inquiries, relatively few demonstrated advanced contextual understanding or integration with broader library services. Response quality varied considerably, with many systems struggling to provide contextually relevant, metadata-rich, or discovery-oriented answers.

A notable finding was the lack of transparency regarding privacy and data practices: only 13% of the chatbots evaluated clearly disclosed privacy information. This gap raises concerns about ethical implementation, user trust, and informed use of AI-enabled services.

The findings highlight opportunities to strengthen semantic capabilities, deepen service integration, and improve transparency to ensure chatbot-enabled services align with institutional values and effectively support diverse user communities.