Authorship in the Age of AI
We live in an age when a sentence that stops you in your tracks might have more than a human hand behind it.
Alan Bennett captured the traditional moment of connection perfectly in The History Boys (2004):
“The best moments in reading are when you come across something — a thought, a feeling, a way of looking at things — which you had thought special and particular to you. Now here it is, set down by someone else, a person you have never met, someone even who is long dead. And it is as if a hand has come out and taken yours.”
If ChatGPT produces a sentence that moves you, it proves what theorists have argued for decades: meaning is created in the act of reading.
The “universal” in literature has always been borrowed, remixed and recycled. Shakespeare did not invent jealousy; he gave Iago words for what every betrayed lover recognises. Mary Shelley did not create the fear of creation; she gave it a form. The magic moment when words “take your hand” does not require a human author, it requires human reading. The connection happens in the reader’s mind, not the writer’s.
Socrates, in Plato’s Phaedrus, distinguished between “living speech” and “dead writing,” arguing that the written word, like seeds left in the sun, has no vitality on its own. Text comes alive only when a reader engages with it.
Readers experience that moment of connection far more often than writers do. For an author, writing is an act of faith. You rarely know when, or whether, your words will find a hand to hold. Sometimes they do, and the reader writes back. Sometimes they misunderstand you entirely. Sometimes they take your work into the world and twist it into something unrecognisable.
The word “author” did not always carry the solitary weight we now attach to it. It comes from the Latin augere, to augment, to grow, to steward. In the pre-modern world, to author was not to conjure from nothing, but to tend to something living: a tradition, a law, a shared story. Medieval poets reassembled episodes, characters and motifs from earlier sources without acknowledgment because borrowing was how one participated in a lineage. To reuse was to honour. The Renaissance crystallised this in the concept of imitatio: true mastery lay in creatively transforming exemplary models. Originality emerged through imitation, not despite it.
Hannah Arendt observed that granting authority to a creator meant granting them a form of custodianship, a responsibility to preserve, interpret and pass on. Modernity replaced this with something narrower. The Enlightenment's investment in the autonomous individual, and the Romantic celebration of the solitary genius, recast authorship as a personal, almost sacred act of creation ex nihilo. Copyright law locked this in, binding authorship to ownership. Creativity became property. Michel Foucault (1969) traced how the figure of the author emerged to regulate creation: to classify, attribute and police discourse. Authorship was less about making things than about controlling them.
AI does not create in the human sense. It reassembles, remixes, extrapolates, the ultimate augmenter, but without intent, without responsibility, without a stake in the tradition it draws from. When an AI generates a poem, a legal argument, or a painting, who is the author? The programmer? The prompter? The thousands of human creators whose work was ingested as training data? Or is the question itself a category error, an attempt to fit an alien process into a human frame?
If authorship was once about stewardship, perhaps AI’s arrival is less a threat than a reminder of what we have lost. In some ways, AI mirrors the old model, it cannot help but be derivative. Yet it lacks the one thing that made traditional authorship meaningful: accountability. A medieval scribe might copy a text imperfectly, but they did so within a web of obligations, to community, patron or God. AI has no such commitments.
Most AI ethics frameworks begin with rights and principles, useful but limited. The Russian philosopher Mikhail Bakhtin (1919) began elsewhere, with authorship and answerability. In his view, responsibility is personal, relational, and inescapably bound to the act of creating. To author something is to be answerable for it. Consider the sculptor who copies a master’s technique or the editor who transforms a manuscript. Each of these is, in Bakhtin’s sense, an author: someone who brings something into presence and answers for what they have made.
Intention matters less than answerability. We do not ask whether a jazz musician intended to create something new when they improvised on a standard. We ask whether the music lives, and whether someone stood behind it.
This reframes what is actually at stake with AI. Who remains answerable for what AI produces? Curation is only authorship when someone is genuinely willing to be held to what they have curated.
There is a Portuguese folk saying: quem conta um conto aumenta um ponto - “whoever tells a tale adds a point”. Every retelling changes the story. Maybe we could say: they who tell a tale, author it a bit. In that light, AI is another layer in the chain of additions, a new way of passing the story along.
And in the act of sharing with us, what you wrote, your words, add to our understanding, and in me commenting, I add a “ point“ to the conversation. Perhaps authorship is just the initiation of a conversation that can become part of a community of thought?