DefinePK

Abstract

The integration of artificial intelligence (AI) into digital media production has transformed the linguistic and cultural landscape of online communication. This study examines the role of AI-assisted meme creation in shaping English language practices, with a focus on lexical innovation, register hybridisation, humour strategies, and global-local dynamics. Drawing on a comparative corpus of 500 memes (250 AI-assisted and 250 human-generated) collected from major social media platforms (2023-2025), the research employs a mixed-method approach combining corpus-based statistical analysis and critical discourse analysis. Findings indicate that AI-assisted memes accelerate the diffusion of new vocabulary, trending hashtags, and euphemistic 'algospeak' while privileging compressed registers marked by abbreviations, emojis, and clipped syntax. In contrast, human-generated memes exhibit a stronger reliance on culturally embedded slang, code-switching, and context-specific humour, such as satire and parody. Humour analysis further reveals a divergence between the algorithmic preference for template-based wordplay and irony, and the human inclination toward sociopolitical satire and pop culture parody.Thus, comparisons highlight AI's tendency toward linguistic homogenisation, but also demonstrate how communities in South Asia particularly in Pakistan recontextualize AI outputs using localized lexicons and hybrid Englishes. The study contributes to debates on mimetics, multimodality, and algorithmic mediation by demonstrating that memes function as laboratories of language change, where algorithmic standardization and human agency intersect. It concludes that while AI fosters novelty and deficiency, human creators ensure the persistence of cultural distinction, linguistic diversity, and digital humour