The growing research literature on the impact of Artificial Intelligence (AI) in higher education is extremely rich in metaphors. This is especially true within the corpus of rapidly proliferating “AI literacy frameworks”, policy documents that outline the skills, competencies, and dispositions needed to engage critically with AI. Within them, conceptual metaphors serve to organize our perceptions and guide our responses to this technological revolution. These metaphors are dynamic, evolving as our understanding changes, keeping in mind that the spread of AI literacy and the interactions between software, students and educators are inherently political.
Based on these premises, our work has both an empirical and a theoretical nature. Through idiographic metaphor analysis, we coded the metaphors included in 18 AI literacy frameworks to map conceptual relationships between AI, students, and teachers. We highlight the dominant metaphors for each actor: AI as [tool-transformer-ubiquitary-artefact-threat]; student as [analyst-citizen-creator]; and teacher as [designer-guide]. We then discuss the natural connections and emerging tensions between these metaphors.
Among these, we note discordance between the metaphorical view of literacy as power and/or adaptation within the frameworks, and too narrow a focus on individual literacy when collective action is needed for coordinating a meaningful response. The framework authors often reference the danger that AI poses to both higher education and representative democracy. However, individual literacy, which is framed too often in the discourse as a panacea for combating the negative impacts of AI, can be viewed from the flip-side as a means for technology companies to consolidate social control and economic gain. Furthermore, we notice how personal responsibility is shifted to the student and teacher away from the AI tool, whose creators hide behind the appearance of technological neutrality. While the technology has become broadly available, its institutionalized adoption and diffusion is not predetermined, but is the outcome of a conscious choice. Note how the simple "exit, voice, loyalty" model first introduced by Hirschman is an excellent heuristic for conceptualizing the current situation.
Finally, we suggest areas for future research that draw on the metaphorical categories outlined in this paper and position literacy as one tool in a multi-aspectual response to AI’s ubiquity in higher education and broader society.
