This article examines how academic discourse is reshaped in higher education through the integration of artificial intelligence (AI) and multimodal design, understood here in the sense of multimodal discourse theory (not multimodal AI models). Grounded in cognitive linguistics, sociocognitive discourse theory and multimodal semiotics, the study analyzes how academic concepts are structured and communicated in AI-enhanced learning environments. It focuses on two Micromodules developed at the University of Osnabrück – Welcome to the AI Jungle and Expedition AI. Micromodules are short multimedia units suitable for Blended Learning that integrate text, visuals, interactivity, and AI-generated feedback within the Stud.IP Learning Management System (LMS). Using a combination of cognitive discourse analysis and multimodal content analysis, the study explores how learners engage with the concepts of learning, argumentation, and autonomy in AI-mediated contexts. Our findings show that learners navigate content using conceptual metaphors like LEARNING IS A JOURNEY, reinforced by modular layout and AI feedback mechanisms. Argumentation is shaped through additive elaboration rather than critical opposition, while autonomy is bounded by interface cues and AI prompts. The study also analyzes how AI systems – specifically tailored to and embedded within the LMS – can participate as semiotic agents, influencing meaning-making through tone, visual presence, and structured interaction. These patterns suggest a shift toward dialogic, hybrid academic discourse in which agency is distributed across human and non-human actors. The article argues that in AI-supported, multimodal learning environments, academic literacy should be reimagined as something created through collaboration between students, educators, and digital tools. It also highlights that developing critical digital literacy is essential for designing curricula that meet the demands of the future.
Dovhaniuk et al. (Tue,) studied this question.