Key points are not available for this paper at this time.
The rapid expansion of Artificial Intelligence (AI) has normalized digital translation in higher education, yet its integration into English as a Foreign Language (EFL) pedagogy remains complicated by an institutional vacuum. Grounded in the Technology Acceptance Model and Cognitive Load Theory, this study examines university students’ multi-tool strategies and critical engagement with Neural Machine Translation (NMT) and Generative AI (GenAI) platforms. Employing a convergent parallel mixed-methods design, quantitative data were gathered from 165 EFL undergraduates (73.7% female, 72.2% sophomores, 61.1% daily active users) via questionnaires, alongside semi-structured interviews with a purposive sub-sample (n=10).The findings reveal an advanced ecosystem characterized by functional differentiation: students strategically leverage specialized NMT engines like DeepL (38.9%) for academic precision but pivot to GenAI tools like ChatGPT (38.9%) for iterative conversational support. While students report high perceived usefulness (98.8%) and ease of use (96.1%) to minimize extraneous cognitive load, they reject passive consumption. Instead, 94.8% actively engage in critical post-editing due to systemic accuracy limitations regarding cultural nuances. Ethically, although 94.7% view automated assistance as an acceptable cognitive scaffold, the absence of clear institutional rules fosters a transparency gap, causing disclosure anxiety. This study dismantles the traditional narrative of student passivity and argues that higher education must move away from obsolete restrictions toward a framework of guided integration and structured translation literacy.
Sulistyani et al. (Fri,) studied this question.