Tracing Adaptive Transfer Across Genres and Media: GenAI ‐Supported ESL Learners' Digital Multimodal Composing and Traditional Writing Practices

Abstract Drawing on the adaptive transfer and the genre knowledge model as a conceptual framework, this research explores how English as a second language (ESL) learners transferred knowledge across genres (narration and argumentation) and media (print‐based writing and video) with the assistance of generative artificial intelligence (GenAI) platforms in the new era of multimodal unified comprehension and generation. The GenAI platforms included text‐to‐text, text‐to‐image, text‐to‐music, text‐to‐video, and text‐conditioned relighting GenAI models, ranging from Microsoft Copilot, Pika, MusicFX, to IC Light. Data were collected from three PhD students' text‐based products, digital multimodal composing (DMC) texts, post‐task interviews, and the first author's observations in a 15‐week extracurricular digital writing program at a public university in Auckland. The results showed that GenAI could lead to ESL learners' transfer of rhetorical and subject matter knowledge across genres. In addition, it was found that these learners transferred this knowledge across media to evoke audience empathy and attention, emphasize the urgency of social issues, and establish authority and credibility. Our findings provide theoretical as well as pedagogical implications for incorporating GenAI into genre‐based writing pedagogy to cultivate students' rhetorical awareness, multimodal GenAI literacy, and writing skill development.