<![CDATA[The rapid development of generative artificial intelligence (Generative AI) technology in the production of Indonesian-language text presents new challenges for maintaining academic integrity and the authenticity of written works. This phenomenon has sparked concerns about the erosion of originality of ideas and the increasingly blurred line between original human thought and algorithmic output. This study aims to formulate comprehensive differentiation indicators to distinguish the authenticity of human authors’ ideas from AI-generated texts based on the Indonesian language. The method used is qualitative descriptive analysis through critical discourse analysis and syntactic analysis of a comparative corpus, consisting of students’ academic papers and AI-generated texts based on similar (prompts). The evaluation focused on the depth of argumentation, cohesion-coherence structure, stylistic dynamics, and syntactic uniqueness. The results indicate that AI-generated texts tend to exhibit rigid sentence patterns, repetitive logical structures, and a lack of constructions reflecting subjective experiential context. In contrast, human writing exhibits rhetorical depth, flexibility in syntactic variation, and originality of ideas rooted in critical reasoning. These findings yield a new analytical framework that higher education institutions can apply to detect and validate the authenticity of academic works in Indonesian more objectively.]]>
Copyrights © 2026
