This study investigates the perceptions, experiences, and evaluative judgments of translation students toward AI-translated idioms and proverbs between Arabic and English. Drawing on a large-scale sample of 3,000 students from public and private Jordanian universities, the research examines how students engage with AI tools in translation tasks that involve culturally embedded figurative language. Using a structured 25-item questionnaire, the study explores six core dimensions: confidence in translating idioms and proverbs, frequency and patterns of AI tool use, perceived accuracy and error detection, attitudes toward AI’s role in translation, preferences for human versus AI translation in idiomatic contexts, and willingness to use AI tools in academic or professional translation. Statistical analysis revealed significant relationships between students’ translation confidence and their ability to identify errors in AI-generated outputs. Participants who frequently encountered idiomatic content and received formal training in translation evaluation demonstrated greater skepticism toward AI’s handling of figurative meaning. While students acknowledged the efficiency and utility of AI tools, a strong preference remained for human translation when dealing with nuanced, culturally rich expressions. The findings support the growing need for AI literacy, critical evaluation skills, and ethical awareness in translator education. Recommendations are provided for educators, curriculum designers, and translation programs to develop balanced, context-aware approaches that integrate human expertise with technological tools.
Haider et al. (Tue,) studied this question.