<![CDATA[Generative AI is widely used in English learning, yet how teachers and learners narrate it as part of their evolving digital identities remains under-researched. This study explores (1) teachers' and students' narratives of AI-mediated English learning, (2) how these narratives reflect digital-identity transformation, and (3) pedagogical principles for humanizing AI. This qualitative narrative inquiry involved five English educators and 15 students from secondary and higher education in Indonesia. Data were collected from interviews, reflective narratives, and anonymized learning artifacts, and analyzed using deductive-inductive narrative thematic analysis. Analysis identified five themes: (1) a shift from apprehension to pedagogical negotiation; (2) reconstruction of learner agency through reflective engagement; (3) reconfiguration of teacher identity from corrector to ethical mediator; (4) emergence of collaborative digital learning identities; and (5) counter-narratives exposing tensions around dependence, authorship, and diminished human interaction, underpinning the proposed Humanized AI-Mediated Digital Learning Identity (HADLI) framework. Generative AI neither humanizes nor dehumanizes English education; its value is co-constructed through teacher facilitation, learner reflection, and ethical transparency, with implications for teacher education, curriculum, and AI policy under Society 5.0. Keywords: humanizing AI, english education, digital learning identity, narrative inquiry, Society 5.0.]]>
