This symposium synthesises four empirical studies exploring the implementation of GenAI-assisted DMC in language education. The findings demonstrate that integrating GenAI into DMC enhances L2 text and multimodal composition and creative translanguaging, enriching teacher feedback and redefining language teachers' identity. Rather than replacing language teachers, this symposium claims that GenAI catalyses pedagogical and professional innovation for language educators. It provides a valuable reference and direction for future empirical research on GenAI-assisted DMC.