<![CDATA[The increasing use of Artificial Intelligence (AI) in academic writing has provided students with support for developing ideas, organizing information, constructing sentences, and improving language. However, the availability of AI assistance does not necessarily mean that the resulting academic text fully demonstrates the characteristics of effective thesis writing. This study examined the characteristics of AI-assisted thesis writing in the Background of the Study sections of five English Study Program students at UIN Sjech M. Djamil Djambek Bukittinggi. The study employed a qualitative approach using qualitative content analysis. Data were collected through document analysis of five students’ Background of the Study sections and were analyzed using a guideline based on Brown’s (2001) writing aspects. The analysis focused on Organization and Vocabulary. Organization was examined through Logical Arrangement of Ideas, Development of Ideas, and Overall Organization, while Vocabulary was examined through Appropriate Word Choice and Contextual Relevance. Researcher coding and second-rater coding were compared, and disagreements were reviewed against the coding criteria before final coding was established. The analysis identified 51 coded instances: 42 related to Organization and 9 related to Vocabulary. Development of Ideas was the most frequent category, with 25 instances, followed by Logical Arrangement with 10 and Overall Organization with 7. Vocabulary findings consisted of 4 Appropriate Word Choice and 5 Contextual Relevance instances. The findings indicate that AI-assisted writing can support recognizable organization and development of ideas, but overall coherence, lexical precision, and naturalness are not automatically guaranteed. The study therefore highlights the continuing importance of critical evaluation and human revision in AI-assisted thesis writing.]]>
