July 16, 2026

Almost every discussion I have surrounding AI-generated art seems to centralize with the question of authorship. Was it the artist who wrote the prompt? The algorithm that generated the image? The engineers who built the model? Or the countless artists whose works became part of the model's training data?

These questions assume that identifying an author is essential to understanding a work of art. However, in his famous essay “The Death of the Author”, Roland Barthes challenged this assumption before AI was even a thought in our brains. For Barthes, the search for an artist's intentions limits interpretation by treating the creator as the final authority on meaning. AI has not disproven Barthes' theory, but in many ways, it has revealed how unwilling we are to abandon the very idea he sought to dismantle.

Rather than asking whether AI can be an author, perhaps we should first ask why authorship remains so central to the way we value art. AI has revived the question of authorship not because Barthes was wrong, or even necessarily right, but because his challenge to traditional ideas of the Author was never fully embraced, or possibly even correctly understood by the public.

When I first encountered Barthes' essay, I saw that it was quite easy to see how the title can be misleading. Barthes is not arguing that artists cease to exist or that their creative labor is unimportant, which is honestly what I originally thought he would do. Instead, he argues that the author's intentions should not serve as the definitive explanation of a work's meaning. The "death" he proposes is not the death of artistic creation but the death of the Author as the ultimate authority over interpretation.

For centuries, Western art has celebrated the artist as a singular genius whose personal experiences unlock the "correct" interpretation of a work. Even in Michelangelo’s commissioned works for the Vatican, people psychoanalyze his own personal struggles with faith to scrutinize how he chose to portray the bodies he painted on the walls and ceilings. Barthes describes this tradition when he writes that "the Author is supposed to feed the book—that is, he pre-exists it, thinks, suffers, lives for it; he maintains with his work the same relation of antecedence a father maintains with his child." In most people’s ideal model, meaning originates with the creator, flows into the artwork, and finally reaches the audience.

Barthes rejects this hierarchy. He points to writers such as Stéphane Mallarmé, whose poetry sought to suppress the authority of the author "for the sake of the writing," shifting attention away from the individual creator and toward the writing itself. Literary movements like this did not erase the creator, but challenged the assumption that the creator alone determined meaning. For Barthes, this shift reflects a larger truth echoed in the writings of many other philosophers: no work is entirely original because every work is composed of ideas, symbols, and language that already exist. As he famously writes, "the text is a tissue of quotations drawn from the innumerable centres of culture." Artists do not create from nothing; they work within the traditions and works that precede them. I know I do with my own artwork. He pushes this idea even further by arguing that "the writer can only imitate a gesture that is always anterior, never original." Creativity is therefore less about inventing something from nothing than arranging inherited cultural material into new relationships. As Ecclesiastes famously says, there is nothing new under the sun.

This idea feels remarkably relevant to conversations surrounding AI, but perhaps not in the way many assume, or at least the way I assumed it would when I first started reading Barthes’ essay. It is important to note that Barthes is not permitting us to declare AI-generated images equivalent to human creativity. Instead, he reminds us that artistic production has never been isolated from culture. Every artist borrows from traditions, visual languages, histories, and images that came before them, made by other humans. Originality has always existed within a network of influence rather than outside it.

If every artwork is already "a tissue of quotations," then AI does not introduce influence into artistic practice; it makes that network of influences more visible. Yet instead of accepting this premise, audiences often react by searching even more intensely for an original creator. AI has not dissolved the myth of originality but has paradoxically strengthened our desire to preserve it.

Most importantly, Barthes argues that meaning does not belong exclusively to the artist. He writes that “Once the Author is gone, the claim to 'decipher' a text becomes quite useless. To give an Author to a text is to impose upon that text a stop clause, to furnish it with a final signification, to close the writing." When we insist on discovering what an artist "really meant," interpretation stops. The artwork becomes something to solve rather than something to experience. This tendency extends beyond literature. Consider Vincent van Gogh. His paintings are frequently interpreted almost exclusively through the lens of his mental illness, as though every brushstroke serves merely as evidence of psychosis. While his biography undoubtedly informs our understanding, reducing the paintings to psychological symptoms narrows their interpretive possibilities. The works become illustrations of Van Gogh's life instead of experiences capable of generating countless meanings for different viewers. This is why Barthes concludes that "the birth of the reader must be at the cost of the death of the Author." Meaning is produced through the encounter between the work and its audience, not recovered from the private intentions of its creator. Even if the creator is an AI program with the work of artists informing its programming, the creator does not technically matter according to Barthes. 

If Barthes hoped criticism would move beyond the search for authorship, AI has had almost the opposite effect. Rather than asking what an artwork communicates or how it produces meaning, conversations surrounding AI almost immediately become investigations into process. It reveals how deeply invested we remain in locating a singular author behind every work of art. In many ways, contemporary art criticism has become almost forensic. Before asking what a work communicates, we investigate minute things like prompts, software, edits, and workflow, hoping to determine exactly who deserves credit through investigation. The artwork itself often waits while we reconstruct its origin story. Ironically, many artists use AI only during the planning stages of a project. An artist might generate compositional studies, experiment with lighting, explore color palettes, or quickly visualize ideas before producing a final painting by hand. In these cases, AI functions much like a sketchbook, a mood board, or another conceptual tool. Yet in my experience, audiences often treat AI-assisted planning differently from other creative aids because AI appears to complicate authorship itself. We become less interested in the finished artwork than in reconstructing how it was made.

Barthes would likely ask whether this is the right place to begin.

The public reaction to AI suggests that Barthes' argument remains deeply controversial. Instead of asking what an artwork communicates, we ask who deserves credit for making it. This instinct reveals something larger than AI, it reveals how difficult it is to separate interpretation from authorship. Even after Barthes argued that criticism should move beyond the Author, we continue to read artworks through the lens of creative ownership. AI has not created this habit; it has re-exposed it and deepened an already complicated theory. If audiences readily accepted Barthes' argument, discussions surrounding AI art might begin with things like emotional meaning or cultural significance. Instead, discussions almost always begin with the question of authorship. So it seems that the public has never accepted that The Author is entirely dead, and I would venture to say that we probably never will.

AI does raise questions that Barthes never attempted to answer. Unlike a paintbrush, AI actively generates images and language in response to human prompts. This complicates discussions of artistic labor, originality, and creative agency in ways that did not exist in 1967. Barthes was not writing about machine-generated content, nor was he attempting to define legal or ethical authorship, but AI also exposes the continued relevance of his theory. If an artwork can involve human judgment, perhaps searching for one definitive author becomes even less useful for interpretation. The artist still makes decisions, but the process itself reveals that creation has always involved more than a solitary genius expressing an entirely original vision. AI does not prove Barthes right or wrong. Instead, it places his theory under new pressure. It asks whether we are capable of separating questions of production from questions of interpretation or whether we inevitably collapse the two.

Maybe, at the end of the day, the most revealing aspect of AI is not the technology itself but the questions it has revived. If, as Barthes argues, every work is already "a tissue of quotations," why does AI suddenly make originality feel so fragile? Perhaps AI has not introduced imitation into art at all, perhaps it has simply made visible what Barthes believed had always been true, the belief that Creation is never isolated from the past. What AI challenges is not originality itself, but our belief that originality depends on identifying a singular creator.

Barthes argued that criticism should move beyond the Author because treating the creator as the source of meaning ultimately limits interpretation. Yet contemporary debates surrounding AI reveal how deeply we continue to rely on authorship as the foundation of artistic value.  This is what makes AI so philosophically significant and morally difficult. Despite decades of literary and artistic theory challenging the authority of the Author, we continue to seek a singular creator before we are willing to engage with a work of art, and the Author is nowhere close to dead.

Works Cited

Barthes, Roland. "The Death of the Author." Image-Music-Text, translated by Stephen Heath, Fontana Press, 1977, pp. 142–48.