Background: The rapid integration of artificial intelligence (AI) into higher education has generated substantial scholarly interest in AI literacy—the competencies required to understand, apply, evaluate, and ethically navigate AI systems. Despite growing publication volume, a systematic cartographic overview of this field remains absent.Objective: This study employs bibliometric analysis to map the intellectual structure, thematic clusters, and developmental trajectories of AI literacy research in higher education from 2020 to 2025.Methods: A corpus of 1,500 peer-reviewed documents was systematically retrieved from four databases (OpenAlex, Scopus, Google Scholar via Publish-or-Perish, and ERIC), merged and deduplicated using the bibliometrix R-package. Analyses included co-authorship networks, keyword co-occurrence, citation analysis, and thematic clustering using VOSviewer.Results: The corpus encompasses 4,201 unique authors across 898 publication sources with 25,755 total citations (mean = 17.17 per document). Annual growth rate reached 207.18%, indicating exponential scholarly momentum. Five AI literacy domains emerge: AI Understanding, Application, Evaluation, Ethics and Responsibility, and Co-Creation, each reflecting distinct pedagogical and epistemological orientations.Conclusions: AI literacy is a rapidly consolidating interdisciplinary field. Recommendations include standardized competency frameworks, ethical dimensions in pedagogy, and cross-disciplinary curriculum design. Findings directly inform policy and instructional practice for preparing higher education graduates for AI-pervasive environments.