Despite the rapid advancements in generative AI technology, the literature lacks a comprehensive exploration of its utilization in Language Assessment within higher education. This mixed-method study explores the intricacies of language assessment typology among Arabic language university teachers, utilizing generative AI. It scrutinizes the correlation between generative AI-driven language assessment methods and the nuanced nature of argumentation within this academic cohort. In quantitative phase, data was collected from 187 university teachers in Arabic-language disciplines using a 5-point Likert scale questionnaire. Exploratory factor analysis was then employed to analyse the data. In the qualitative phase, interviews were conducted with 16 university teachers selected from the quantitative phase and their document tasks were analysed thematically. The results revealed four language assessment typologies among university teachers, corresponding to different types of argumentation: formative diagnostic, summative, and peer assessment. Formative assessment was linked to enriching interactive classroom discussions, diversification of learning activities, uncovering evidence of students’ comprehension, and fulfilling various cognitive functions. However, it’s imperative to develop a questioning model to optimize its effectiveness. Diagnostic assessment was associated with argumentation that the use of generative AI facilitated engagement in higher- level language activities, diagnosing and refining students’ ideas, and justifying the validity of any explanation. Summative assessment was associated with supporting scientific reasoning, enhancing argumentation skills, and gauging overall language proficiency. However, establishing clear argumentation guidelines is imperative. Peer assessment was linked to facilitating peer feedback, fostering proficiency in generative artificial intelligence, promoting collaboration and communication among students, and enhancing both learning and assessment processes. However, effective feedback for both the provider and recipient is crucial.