Asia A. AlheetyMethaq Kh. MararHusam J. Mohammed2026-01-132026-01-132025Asia A. Alheety,Methaq Kh. Marar,Husam J. Mohammed. (2025). A Comparative Analysis of Lexical Richness in AI-Generated Vs. Human-Authored Literary Articles. International Conference of Language & Education (ILEC2025), 749–769. https://sites.google.com/view/ilec2025/home3120-3809https://sites.google.com/view/ilec2025/homehttps://oarep.usim.edu.my/handle/123456789/28502International Conference of Language & Education (iLEC2025): Empowering Language and Education to Develop Madani Generation in the Era of Artificial Intelligence (AI) 17 & 18 November 2025 Organized by Universiti Sains Islam MalaysiaThe increasing prevalence of AI-generated texts raises critical questions about their linguistic authenticity compared to human-authored content. This study presents a comprehensive comparative analysis of lexical richness between AI-generated and human-authored literary articles using Halliday's systemic functional linguistics (SFL) framework. A corpus of 10 literary articles was selected (qualitative method, due to content analysis), comprising 5 human-authored texts and 5 AI-generated counterparts matched for length and genre. The analysis employed multiple lexical richness metrics including Type-Token Ratio (TTR), Measure of Textual Lexical Diversity (MTLD), lexical density, lexical sophistication, and Halliday's three meta-functions (ideational, interpersonal, and textual), (quantitative method, due to metrics analysis). Depending on that this work will be mixed mode method. Using purposive sampling, texts were selected based on comparable length, genre, and publication standards. Results reveal significant differences between human and AI-generated texts across several dimensions. Human-authored texts demonstrated higher variability in lexical choices (TTR: 0.683 vs 0.671), greater sentence complexity (25.12 vs 22.29 words per sentence), and significantly higher ideational density (0.026 vs 0.010) and interpersonal density (0.031 vs 0.010). Conversely, AI-generated texts exhibited higher lexical density (0.771 vs 0.696), greater lexical sophistication (0.596 vs 0.436), and longer average word length (6.95 vs 5.94 characters). These findings suggest that while AI systems excel at producing grammatically sophisticated and lexically dense texts, they lack the functional variety and interpersonal engagement characteristic of authentic human communication. These findings have significant implications for AI development, automated writing assessment, and understanding the fundamental differences between human and artificial linguistic production.en-USlexical richnessartificial intelligencecomputational linguisticssystemic functional linguisticscorpus analysistext generation.text::conference output::conference proceedings::conference paper749769