This work is a case study in understanding the argumentative functions of metaphor within literary discourse by means of novel integration of AI techniques with traditional literary analysis, specifically applied to the poetry of Ibn al-Jayyab al-Gharnati (d. 815H), one of the most celebrated classical Arabic poets. The nature of metaphors in Arabic poetry has often been historically valued for aesthetic or symbolic reasons, with limited research investigating their argumentative and persuasive use. Using artificial intelligence and natural language processing (NLP) algorithms, this study implements a systematic approach to extract metaphorical expressions from the poetry of Ibn al-Jayyab and analyses their contribution in creating argumentative accountability as well as rhetorical effectiveness. It employs a mixed-methods design that reveals not only the frequency, but also the functional distribution of metaphors in Ibn al-Jayyab’s poetic corpus. This first step qualitative analysis looks at the context and purpose of single metaphors to reveal its argumentative goals for the poet. While Mazumdar’s work is primarily qualitative and relies on close readings of texts to interpret metaphors, the quantitative analysis uses statistical methods to find trends in metaphor use—how some metaphoric frames (such as light versus darkness or journey versus path) become repeated motifs that emphasize important philosophical and ethical points. The results show that Ibn al-Jayyab’s metaphors are not ornamental, but rather intentional rhetorical devices that reinforce the poet’s claims and push readers towards certain interpretations. For instance, light and darkness symbolize knowledge and ignorance respectively, which creates an orientation of the audience ‘identities and values. Additionally, the study shows how AI tools help design in revealing metaphorical tabular structures to readers that might otherwise be obscured to human analysis alone, confirming the valued place of AI as augmenting literary studies.