The Transparent Ghostwriter: Artificial Intelligence as a Legitimate Support in Scientific Writing

The use of artificial intelligence in writing processes is often met with suspicion, particularly in academic contexts. Yet, the involvement of external contributors in text production (ghostwriters, anonymous proofreaders, human editors) has long been widely tolerated in other domains, without explicit requirements for transparency or formal disclosure.This essay seeks to address the following question: why is the intervention of a human ghostwriter considered acceptable within the publishing world, while the declared use of artificial intelligence continues to generate strong distrust? The article argues that artificial intelligence, when employed as a support tool and used transparently, does not raise ethical issues different from those associated with traditional editorial collaboration. On the contrary, AI can be interpreted as a transparent ghostwriter, capable of exposing structural hypocrisies embedded in systems of textual and editorial production.Through an analysis of the actual writing process within a scientific research center, the paper shows how AI intervenes in functions that have long been delegated to external actors, human and generally invisible. The analysis focuses on editorial dynamics and evaluation mechanisms that have contributed to making the integration of artificial intelligence into scientific writing problematic.