The present study aimed to validate and evaluate the psychometric properties of the Persian version of the Artificial Intelligence (AI) Literacy Scale in Second Language Writing among student teachers. Adopting a descriptive-survey design, this research falls within the scope of instrument development studies. The statistical population comprised all student teachers at Farhangian University during the 2025–2026 academic year, of whom 828 individuals (485 females and 343 males) were selected via convenience sampling. To ensure construct validity, the total sample was randomly split into two subsamples. Exploratory Factor Analysis (EFA) conducted on the first subsample extracted a four-factor structure—comprising Understanding, Usage, Evaluation, and Ethics—which accounted for 56.15% of the total variance. Subsequently, Confirmatory Factor Analysis (CFA) on the second subsample confirmed the robust fit of the extracted model and the hierarchical structure of the scale. The instrument demonstrated high internal consistency with a Cronbach’s alpha of 0.93 and composite reliability (CR) values exceeding 0.80. Furthermore, criterion-related validity was established through a significant positive correlation (r = 0.73) with the "Attitude toward AI Use in L2 Learning" construct. Findings revealed that while the students' overall AI literacy was above average, the 'Ethics' component yielded the highest mean score, whereas 'Evaluation' received the lowest. Overall, this scale is a valid and reliable tool for assessing AI literacy in L2 writing within the Iranian higher education context and can serve as a framework for designing teacher training programs focused on critical thinking and ethical engagement with emerging technologies.