The Ghost in the Dissertation: Ethical Hauntings of AI in the PhD Journey

Abstract Advances in generative artificial intelligence (AI) have introduced new possibilities and new anxieties into doctoral scholarship. As large language models (LLMs) become increasingly embedded in writing, analysis, literature review, and methodological design, doctoral researchers encounter a constellation of ethical “hauntings”: the invisible presence of AI that complicates authorship, originality, mentorship, and scholarly identity. This article interrogates the ethical, epistemological, and pedagogical implications of AI’s presence in the PhD journey. Through interdisciplinary literature, conceptual analysis, and qualitative reports from emerging research, the paper examines three dimensions of AI “ghostliness”: (1) the spectral author—ambiguities in academic integrity and authorship; (2) the haunted method shifts in epistemic labour and methodological design; and (3) the disembodied mentor changes in doctoral supervision, labour, and learning. The article concludes by offering a framework for ethical AI companionship that preserves scholarly agency while acknowledging AI as a legitimate though carefully bounded research aid.