<![CDATA[Background: Generative AI has unsettled conventional authorship in Indonesian higher education by blurring boundaries between legitimate writing support, linguistic development, and academic misconduct. Objective: This study examines how authorship after generative AI is constructed through assessment design, academic integrity governance, and linguistic evidence in Indonesian higher education. Method: Using qualitative corpus-based document analysis, this study analyses 17 publicly accessible policy, pedagogical, contextual, and scholarly documents through coding categories on accountability, assessment orientation, and language-based authorship judgement. Results: Findings show that human-accountability centred authorship dominates corpus documents, indicating that students remain responsible for claims, sources, reasoning, and final submissions even when AI assistance is permitted. Findings also reveal that assessment responses increasingly privilege process-based evidence, pedagogical guidance, oral explanation, prompt documentation, and source verification over detector-only control. Implication: Linguistic evidence is positioned as supportive rather than conclusive, because polished academic English may reflect EFL development, translation, feedback uptake, or AI-mediated revision. Novelty: This study offers a novel assessment-ecology perspective by integrating authorship accountability, integrity design, and contextual linguistic judgement as a defensible framework for evaluating AI-assisted academic writing in multilingual higher education]]>
Copyrights © 2026
