Empowering Linguistically Responsive Teachers with Text-to-Video Generative AI as a Multimodal Intelligent Classroom Assistant

• In Event: Language and Literacy Development and Decolonization through AI and Robotics

Abstract

This study explores how Text-to-Video Generative AI tools (T2V-GenAI), VEED.io and CapCut, can support culturally and linguistically responsive teaching in multilingual classrooms as a Multimodal Intelligent Classroom Assistant. Using a collaborative action research study, researchers guided preservice and in-service teachers to create multilingual and multimodal children’s books and read-aloud videos by typing scripts and incorporating voice-over and captions in multiple languages, customizable avatars, and autogenerated illustrations to provide translanguaging spaces. This approach showcased how teachers can move beyond static content delivery toward a more differentiated and inclusive curriculum and instruction by incorporating T2V-GenAI tools. Despite technological limitations, findings demonstrate that T2V-GenAI tools can act as effective classroom assistants, promoting equitable access to content and affirming students’ cultural identities in classroom settings.

Authors

• Matthew Nyaaba, University of Georgia

• SungEun Min, Kutztown University of Pennsylvania

• Gayoung Choi, University of Nebraska - Lincoln