Background: Academic library discovery systems increasingly mediate access to scholarly information through ranking, metadata matching, filtering, database coverage, availability signals, and recommender functions. These algorithmic arrangements can shape what users notice, cite, and trust. Objective: This paper examines how algorithmic bias, user search behaviour, information trust, and digital information literacy intersect in academic discovery environments. Methodology: Since no local institutional dataset was provided, the paper uses a secondary analysis of online primary empirical evidence from three verifiable sources: Project Information Literacy's student survey of 2,318 respondents across six U.S. campuses; a published Summon query/click-log analysis based on anonymized searches by millions of users at hundreds of libraries; and an open representative German online survey on search-engine literacy and trust involving 2,012 respondents. Results: The data show high dependence on scholarly databases and OPACs but limited engagement with librarians. In the Project Information Literacy survey, 84% of students used scholarly research databases and 78% used OPACs for course-related research, while only 20% consulted librarians about assignments. Summon logs show that users generally began with simple search, with fewer than 6% of queries using fielded search and only slightly more than 2% using advanced operators, although more than 40% used at least one filter. Search-engine literacy data show that 79% knew paid ads exist, but only 29.2% knew how ads differ from organic results, and only 8.9% knew the term SEO. Implications: Libraries need discovery-literacy instruction that connects searching, ranking, metadata quality, database coverage, and calibrated trust.

Keywords: algorithmic bias; academic libraries; discovery systems; information trust; search behaviour; digital information literacy; information retrieval