SILENCE IN AI TEXTS: INVESTIGATING THE LOSS OF INTERPERSONAL MEANING IN GENAI-PRODUCED ACADEMIC PASSAGES

The growing use of Generative Artificial Intelligence (GenAI) in English for Academic Purposes (EAP) teaching has profoundly changed the way teachers and students write and read academic texts. Nevertheless, behind the fluency and accuracy of the passages generated by AI-machines, lies a subtle silence viz. the disappearance of the human voice that gives writing its warmth, stance, and dialogue. The present study examines how GenAI-produces academic passages, expresses, or fails to express, interpersonal meaning. Grounded in the appraisal theory of the Systemic Functional Linguistics (SFL) framework, the study compares AI-produced and human-written texts on two distinct EAP topics. It uses a combination of corpus analysis and qualitative discourse examination to scrutinize how these texts express attitude, hedge claims, or invite reader alignment, and authorial presence. The findings revealed that GenAI texts underrepresent the subtle interpersonal cues that give writing its persuasive, dialogic, and ethical texture, though they excel in both grammatical accuracy, and lexical variety. Moreover,they sound too neutral, use fewer engagement signals, and lack guiding phrases that connect ideas. Such loss of interpersonal nuance in GenAI texts causes them to sound correct but detached from the intellectual dialogue that academic literacy requires.