Machine translation and post-editing in translator training: a systematic review of integration models, challenges, and pedagogical implications

The integration of machine translation (MT) and post-editing (PE) into translator education has become increasingly important amid rapid advances in neural MT and large language models. While research has expanded in recent years, evidence on how MTPE is embedded in curricula, how students engage with it, and what challenges persist remains scattered. This article systematically reviews peer-reviewed studies on MTPE in translator training. Using PRISMA-based selection, studies were analyzed across four dimensions: (1) temporal developments, methodological orientations, and language pair configurations; (2) engagement with specific text types, learner proficiency levels, and pedagogical objectives; (3) integration of MTPE into curriculum design and competence development; and (4) persistent challenges and gaps in MTPE pedagogy. Findings show a steady expansion of MTPE scholarship, with methodological diversification but strong English-centric bias. Research mainly evaluates training effectiveness, learner perceptions, and technological or curricular innovations. Pedagogical models range from stand-alone MTPE courses to integrated modules and hybrid learning, increasingly targeting skills in error identification, strategic post-editing, and reflective practice. Yet unresolved issues include overemphasis on technical skills, limited attention to cognitive and affective dimensions, and inequities in tool access across languages. This research argues for more holistic, learner-centered, and context-sensitive integration models that can prepare future translators for AI-augmented professional environments. The findings also provide a research agenda for advancing MTPE pedagogy through methodological triangulation, curricular coherence, and equitable access to technological resources.