AI and Literary Creativity: Who is the Author?

GPT Writes a Novel

Language models generate texts indistinguishable from those written by humans in short formats. Longer contexts are more complex: coherence breaks down, “hallucinations” increase. But GPT-4 is capable of creating convincing stories, poetry, and dialogues.

This raises philosophical questions. What makes a text “literary”? Beauty of form? Depth of meaning? Consequences for the author? Reader’s experience? If AI creates a text that evokes a genuine emotional response in the reader—is that “literature”?

The “romantic author” is a concept of the 19th century: a text has a single genius-author, the source of meaning and authenticity. Foucault (“What Is an Author?”) challenged this: the “author” is a function of discourse, not the biological source of the text. AI-generated literature continues this question.

Practice: AI as a Writer’s Tool

Many professional writers already use AI as a tool: generating options, overcoming “writer’s block,” checking coherence, speeding up drafts. It is not a “replacement”—it is a tool, like a word processor.

But the boundaries are blurring. “Ghostwriting” has always existed. Is “AI-assisted writing” its new form? Or is it a qualitatively different phenomenon? If a human writer edits 10% of the AI-generated text—who is the author?

Copyright: courts in different countries provide different answers. In the US—AI cannot be an author. In the EU—there is a debate. This is not only a legal question: it is a question about the nature of creativity and its value.

Question for reflection: You have used AI to create texts in your work. How do you draw your personal ethical line between “AI as a tool” and “AI as author”? What preserves value in your personal text creation?