Speaker
Description
The rapid adoption of Generative Artificial Intelligence (Gen AI) in second language (L2) higher education offers new affordances for multilingual interaction, multimodal expression, and individualized feedback. At the same time, growing concerns around academic integrity, erosion of critical thinking, equity, and the reproduction of cultural and linguistic bias remain underexplored. Existing technology-integration frameworks such as Substitution Augmentation, Modification Redefinition (SAMR) (Puentedura, 2013) have been criticized for prioritizing tools over learner agency (Blundell et al., 2022), while recent pedagogical approaches foregrounding translanguaging have not fully addressed assessment, reflection, or multimodal task design in AI-supported learning (Donley, 2024).
Responding to these gaps, this conceptual article proposes an inclusive framework for Gen AI–mediated task design grounded in translanguaging pedagogy and social semiotic theory. Drawing on current literature, the paper examines how Gen AI tasks designed through this lens can support three interrelated dimensions of learning: (1) critical thinking, by engaging learners in the analysis, critique, and revision of AI-generated content; (2) translanguaging and multimodality, by enabling learners to mobilize diverse linguistic and semiotic resources; and (3) equity and representation, by addressing issues of bias, access, and inclusive participation. Gen AI is thus positioned not as a replacement for instruction, but as a complementary pedagogical tool that supports reflective, collaborative, and culturally responsive ESL learning. The article concludes by outlining implications for ESL-oriented pedagogy and identifying directions for future research, particularly the need to move beyond single-text literacy tasks toward multimodal, skills-integrated, and reflective learning experiences.
