![[ Visit Client Website ]](images/banner.gif)
The proliferation of Artificial Intelligence (AI) technologies throughout the U.S. higher education landscape has gained significant momentum. However, AI literacy development among higher education students has not kept pace with the rapid expansion of AI use. The purpose of this study is to systematically map trends and characteristics of peer-reviewed research on AI literacy among higher education students in the United States (U.S.), including publication trends, methodological approaches and sample characteristics; conceptualizations of AI and AI literacy; levels and measurement of AI literacy; factors associated with AI literacy; interventions designed to enhance AI literacy and their outcomes; and recommendations for future research, policy, and practice.
Methods:
This scoping review adopts the Joanna Briggs Institute six-stage framework for scoping reviews: (1) research question formulation, (2) study identification, (3) selection, (4) data charting, (5) synthesis and reporting, and (6) expert consultation. The comprehensive literature search was executed in May 2025 across five electronic databases, including Academic Search Premier, APA PsycInfo, Education Source Ultimate, ERIC (ProQuest), and Web of Science. The screening process involved two independent reviewers who evaluated titles, abstracts, and full texts based on pre-defined inclusion and exclusion criteria, with any conflicts resolved through consensus. A total of ten studies were ultimately included for the scoping review. Covidence was utilized as a systematic review tool to screen and extract data from the selected literature.
Results:
The majority of scoped studies were published between 2024 and 2025, underscoring the field’s nascent and underexplored nature. The studies were predominantly situated in public institutions and education-focused journals, with limited representation across geographic regions, private institutions, academic disciplines, and diverse student populations, including variations in gender, race/ethnicity, and generational status. Research designs primarily employed mixed-methods approaches; however, AI literacy measurement remains underdeveloped and relies heavily on self-reported confidence rather than standardized or objective assessments. Conceptually, students tend to perceive AI as a collaborative tool, whereas researchers emphasize technical components, revealing a gap between user perceptions and scholarly frameworks. Interventions ranged from brief lectures to longitudinal curricula and demonstrated potential to enhance student self-efficacy, conceptual understanding, and technical skills, particularly regarding AI ethics, output evaluation, and advanced prompting strategies.
Conclusions and Implications:
The findings provide important implications for social work research and education. The limited exploration of AI literacy in higher education suggests a critical need to examine how social work students and practitioners understand and engage with AI technologies. Social work research should investigate the ethical, relational, and sociocultural dimensions of AI use, including equity, bias, and client-centered applications. Developing standardized and context-sensitive measures of AI literacy is essential to inform curriculum design and professional training. Furthermore, interventions that integrate AI tools with human-centered social work practices may enhance students’ technical competencies while maintaining the field’s emphasis on ethical decision-making, client advocacy, and culturally responsive care.