Security and Privacy Challenges in AI-Powered Library Recommender Systems: A Systematic Literature Review

Abstract

As AI tools become more embedded in library systems, concerns around privacy, security, and ethics are rising. This study presents a systematic literature review examining AI-powered recommender systems within library contexts, with a focus on their applications, associated risks, and governance. The literature reviewed highlights key security and privacy risks, including data poisoning, inference attacks, and algorithmic manipulation, which threaten core library values such as intellectual freedom and patron confidentiality. Existing policy frameworks, such as the GDPR and the AI Act, offer guidance on fairness and transparency, but a lack of library-specific governance standards persists. A balanced security approach is proposed, consisting of technological safeguards, AI awareness training among staff and patrons, and the creation of library-specific standards.