Is the Use of Generative AI in Academic Writing Legitimate? Rethinking Authorship Through Three Forms of AI Assistance

The rapid growth of generative artificial intelligence (GenAI) has presented new opportunities for academic writing. At the same time, it has raised difficult ethical questions regarding authorship, originality, and responsibility in scholarly work. The current discussions tend to categorize the use of AI in academic writing as either acceptable or unethical, thus establishing a binary framework that overlooks the differences through which academics make use of this technology. This paper argues that the legitimacy of AI-assisted academic writing should instead be understood as a spectrum of human–AI involvement. Adopting a conceptual analytical approach, the study synthesizes recent scholarship and policy discussions on generative AI in higher education and scholarly publishing. Based on this synthesis, the paper proposes a three-part framework that distinguishes among AI as linguistic support, AI as cognitive partner, and AI as ghostwriter. These categories represent increasing levels of AI participation in the writing process and provide a basis for differentiating between supportive assistance and practices that challenge conventional understandings of authorship. This analysis calls attention to the fact that AI used as a form of language support fits largely within existing editorial practices. The same cannot be said of AI as a cognitive partner, which occupies an intermediary and complex role requiring significant human guidance and accountability. On the contrary, when viewed as a ghostwriter, AI entails severe ethical problems and is fraught with significant concerns related to the transfer of meaningful intellectual work to machines. Overall, this paper makes a significant contribution to the discussion regarding the concept of AI-assisted writing by elaborating a refined classification of the forms of such technology in use.

کلیدواژه ها:

Generative artificial intelligence ، academic writing ، authorship ، AI-assisted writing ، academic integrity ، human–AI collaboration