https://doi.org/10.67147/literariness.v1i3.032
The Inherited Voice: AI, Authorship, and the Irony of Machine Recognition
PARTHI CHAKRABORTY
Darrang College, Tezpur, Assam
parthichakraborty@gmail.com
Abstract: Not long ago, I submitted an article I had written entirely on my own to an agency. Their software flagged it as 75% AI-generated. Another machine had decided that my writing was not human enough. I sat with that for a while. Barthes declared the death of the author, and Foucault questioned the very function of authorship. Yet neither could have imagined a moment when a writer completes a deeply personal piece of work only to be informed by an algorithm that it is probably not hers at all.
Artificial intelligence has spent years learning from human writing. It has absorbed the ways we construct sentences, create rhythm, and express emotion. It has learned to sound like us. This is not creativity so much as inheritance. And when inheritance occurs without consent, it begins to resemble loss. This paper explores the nature of that loss—not the loss of employment or relevance, but the quieter loss of a voice one believed to be uniquely one’s own. Drawing on the theoretical perspectives of Roland Barthes and Michel Foucault, and beginning from the experience of a falsely flagged article, the study investigates what human writing possesses that machine-generated writing can only imitate. It ultimately reflects on authorship, originality, and the changing boundaries between human expression and algorithmic reproduction in the age of artificial intelligence.
Keywords: Authorship, Artificial Intelligence, Barthes, Foucault, Creative Identity, AI Detection, Human Voice
Read Full Manuscript PDF