Ethical determinants of perceived ethical legitimacy in AI-driven information services: Evidence from Thai academic libraries

The rapid integration of artificial intelligence (AI) into information services has raised critical questions about ethics, privacy, and user trust. This study explores the ethical and privacy concerns surrounding AI-based information services. It investigates their implications for perceived ethical legitimacy, which is conceptualized as a precursor to user trust in AI-based information services. A quantitative research design was employed, utilizing a structured survey distributed to 278 library users in Thai universities. The instrument, validated through expert review and reliability analysis (Cronbach’s alpha = 0.870), measured perceptions of 10 ethical constructs, including data responsibility and privacy, fairness, security, transparency, and accountability. Data were analyzed using correlation and multiple regression techniques. The results revealed that while users overwhelmingly perceived AI in libraries as beneficial (mean score = 4.34/5), perceptions of ethical legitimacy were strongly associated with ethical safeguards, which may influence trust formation. Specifically, data responsibility & privacy, fairness, and ethics & regulations emerged as significant predictors of overall perceptions of AI ethics, collectively explaining 26.6% of the variance. Other constructs, such as security, accountability, and trust, while valued, did not demonstrate unique predictive power within the model. The findings underscore that ethics are not peripheral but central to technology acceptance in information services. This study contributes to theory by extending technology adoption frameworks to include ethical dimensions and to practice by offering evidence-based guidance for policymakers and library administrators to prioritize privacy, fairness, and regulatory clarity in AI deployment. Although limited to a single national context and cross-sectional design, these findings provide a foundation for future comparative and longitudinal research. Ultimately, the study highlights that sustainable adoption of AI in academic libraries depends not only on technical innovation but also on embedding robust ethical practices that foster user trust and confidence.