NOT REPLACED, BUT REDEFINED

Abstract

As machine translation (MT) grows increasingly fluent, the human translator is not being replaced but redefined. This study explores that redefinition within English as a Medium of Instruction (EMI), where content and language learning intersect. In EMI classrooms, lecturers and students often act as de facto translators -negotiating meaning across languages using AI-powered translation tools. Adopting a theoretical descriptive methodology, the paper aims to investigate how MT reshapes the human translator's role from sole author to critical curator. Findings revealed that when used collaboratively, MT enhances cross-linguistic comprehension, accelerates feedback, and allows human agents to focus on cultural nuance, pragmatic inference, and ethical judgment—tasks machines cannot yet own. However, risks persist: algorithmic bias, over-reliance on automated fluency, and the erosion of students' own translation skills. The study proposes a pragmatic, human-centered model: machine translation as a co-pilot, not a replacement. In EMI contexts, this means transforming translation from a hidden necessity into a visible, teachable competence. The human translator of the AI era becomes a pedagogic mediator-guiding, correcting, and culturally mediating between algorithmic outputs and learning goals. Ultimately, AI tools complement rather than replace human instruction, bridging pedagogy with digital transformation while preserving ethical responsibility