Navigating new Frontiers: university-level English language teachers' perceptions of AI-generated grammar exercises in language learning

Purpose Artificial intelligence (AI) has recently been integrated into language teaching and assessment, revolutionizing the evaluation of language skills. Therefore, this study aims to explore the perceptions of university English language teachers towards AI-generated grammar assessments. It identifies the differences between AI-generated and human-made tests, reveals the potential benefits and limitations of AI-generated items and proposes practical strategies to overcome these limitations. Design/methodology/approach The data were collected from 12 experienced English as a Foreign Language (EFL) teachers at Saudi Electronic University, using a mixed-methods approach that comprised a test evaluation checklist and semi-structured interviews. The checklist, adapted from Brown (2004) and Brown (2005), evaluated the test items in terms of validity, reliability, contextuality, involvement and quality. Findings The results revealed that the human-made test excelled in validity, contextuality and involvement, while the AI-generated test demonstrated strengths in reliability and clarity. The thematic analysis of the interviews’ responses identified three benefit-related themes: time-saving, question variety and personalization; five challenge-related themes: ambiguity, poor prompts, over-reliance, ethical concerns and errors; and two suggested strategies: monitoring and human editing. Originality/value While using AI has promising potential, its effective implementation in language classrooms requires careful planning. Thus, in light of the findings, the study proposes practical strategies to bring out the strengths of AI while mitigating its limitations.