The advent of large language models (LLMs) has fundamentally reshaped the translation paradigm, transitioning from traditional human-centered workflows to a human-AI collaborative model. Advanced LLMs now support translators throughout the entire translation pipeline, encompassing pre-translation planning, drafting, post-editing, and evaluation. These generative AI systems augment human cognitive capabilities, rendering the translation process increasingly AI-driven. Accordingly, the translator’s role is redefined—from that of a linguistic executor to a strategic mediator who orchestrates, refines, and critically evaluates machine-generated content. This transformation not only improves efficiency and quality but also reconfigures the core professional competencies required, emphasizing AI literacy, prompt engineering, and collaborative problem-solving. As a result, translation education must adapt to equip future professionals with the skills necessary to thrive within emerging human–machine translation ecosystems.