The emergence of large language models capable of producing fluent poetry, memoir, and fiction has renewed long-standing critical debates about authorship, creativity, and aesthetic agency. This paper examines recent literary experiments in human-AI co-creation to argue that generative AI does not simply revive or resolve the poststructuralist critique of the author associated with Barthes and Foucault; it stages that critique in a new, materially specific form.
Drawing on Colella’s analysis of the enlarged paratexts surrounding AI-generated literature and Bajohr’s model of graduated authorship, from primary human composition to quaternary prompt-based generation, the paper argues that the human figures who curate, prompt, and edit AI output have not disappeared from these works but have instead relocated their authorial presence into surrounding textual apparatus: prefaces, afterwords, footnotes, and public commentary that explain, validate, and claim responsibility for the machine’s language. Benjamin’s earlier account of aura and mechanical reproduction offers a useful historical parallel, suggesting that each new reproductive technology prompts a similar renegotiation of where authenticity and authorial presence are felt to reside.
The paper concludes that AI-generated literature, far from eliminating the author, has intensified the visibility of authorial labor by displacing it from the sentence to the surrounding frame, making paratext, rather than prose style alone, the primary site where contemporary readers are asked to locate a human creative presence.