When did we stop calling it ghostwriting?
John F. Kennedy won the Pulitzer Prize for Biography in 1957 for Profiles in Courage. His name was on the cover, his photograph on the jacket, and the book helped cement his reputation as a thoughtful statesman on his way to the presidency. His speechwriter, Ted Sorensen, later acknowledged that his own role in the book's creation was substantial enough that the question of authorship followed both men for decades.
This is not a scandal. It is a footnote. Historians mention it, biographers note it, and nobody is writing angry op-eds demanding the Pulitzer be returned. We have, collectively, decided that this is fine.
Which is interesting, because if you told someone that a political figure used a sophisticated tool to produce a book he didn't actually write, then put his name on it and won a major literary prize — well, in 2026, that sentence would start a very different conversation.
The industry of invisible authors
The ghostwriting industry is enormous and almost entirely uncontroversial. Celebrity memoirs are overwhelmingly written by professional ghostwriters whose names sometimes appear in small print on the acknowledgments page and sometimes don't appear at all. When you read a memoir by a professional athlete, a reality television star, or a retired general, you are almost certainly reading the work of someone you have never heard of, filtered through interviews and conversations into a first-person narrative designed to sound like someone who does not write for a living.
Nobody minds. We understand the deal. The celebrity provides the experiences and the name; the writer provides the sentences. It is a transaction we have all tacitly agreed to accept.
Corporate ghostwriting is even more pervasive and even less discussed. The "thought leadership" articles attributed to Fortune 500 CEOs are written by communications teams or hired agencies. Speeches delivered at conferences are drafted by speechwriters. Those heartfelt LinkedIn posts from founders with millions of followers — the ones with the short paragraphs and the hard-won lessons and the single-sentence kicker at the end? Many are crafted by content strategists who have studied the founder's voice and can reproduce it on command. The founder might review the draft, suggest a few edits, or simply approve it with a thumbs-up emoji. The post goes out under their name. Their followers engage with it as a personal expression of the founder's authentic ideas.
This, too, is fine. Everyone involved understands what is happening, and everyone has quietly agreed not to find it troubling.
Authorship as a polite fiction
Academia presents an even stranger case. A professor's name appears on a paper because they run the lab, secured the funding, and provided general direction — even when graduate students conducted the research and wrote the manuscript. In some fields, the head of a department is listed as a co-author on dozens of papers per year, a physical impossibility if "authoring" requires having written any of the words.
The International Committee of Medical Journal Editors established authorship criteria decades ago. The rules state that to qualify as an author, you must have made substantial contributions to the work's conception or design. You must have drafted the work or revised it critically. You must have approved the final version. You must agree to be accountable for all aspects of the work.
In practice, many listed authors satisfy perhaps one of these criteria and sometimes none. The system runs on prestige, hierarchy, and convention. It is, to use a technical term, ghostwriting with extra steps.
So here we are, in a culture that has comfortably accommodated ghostwriting in politics, entertainment, business, and academia for as long as any of these institutions have existed. We have developed elaborate, unspoken agreements about when it is acceptable for the person whose name is on the work to not have written the work. The rules are never stated explicitly because stating them would require admitting they exist, and the whole arrangement works precisely because nobody examines it too closely.
And then a machine enters the picture, and suddenly everyone has very strong opinions about authorship.
The line nobody drew
The interesting thing about the outrage over AI-generated writing is not that it exists — the discomfort is real, and worth taking seriously — but that it requires a line to exist somewhere between acceptable and unacceptable not-writing-it-yourself, and nobody can quite agree on where that line is or when it was drawn.
Consider a spectrum. At one end, a person sits alone and writes every word. At the other end, a person types a prompt and publishes the output without reading it. In between, there is a vast and densely populated middle ground: the executive who gives notes to a speechwriter; the celebrity who talks for twenty hours into a recorder while a ghostwriter shapes those ramblings into a memoir; the academic who outlines a paper and hands it to a graduate student; the blogger who drafts something rough and lets an editor rewrite half of it; the founder who approves a LinkedIn post written entirely by a content agency.
All of these are common. All of them are accepted. And in every case, the named author did not write the sentences — someone or something else converted their ideas, or their experiences, or merely their name recognition, into finished prose.
Now add: the writer who uses an AI to produce a first draft from detailed notes, then spends an afternoon restructuring, rewriting, and reworking it until the piece says what they meant. Where does this fall on the spectrum? It involves more direct engagement with the text than the celebrity memoir. It involves more creative judgment than the CEO's agency-written thought leadership piece. And yet it is the one that draws accusations of fraud.
What "authentic" means, and doesn't
The most common objection is authenticity — the idea that AI-generated writing does not represent the genuine thoughts and experiences of the person whose name appears on it. This is a reasonable standard, but an oddly selective one. A ghostwritten memoir is considered authentic because the experiences are real, even though the sentences belong to someone the reader has never heard of. The celebrity provides the life; the writer provides the language. The arrangement is well understood and entirely uncontroversial.
When an executive uses an AI to draft an article from their own ideas and notes, the structural relationship is similar — one party provides the raw material, another converts it into polished prose. The difference people point to is that a human ghostwriter "understands" the ideas, while an AI merely processes them. This may be true, but it is a claim about consciousness rather than about writing. It moves the question from "who wrote the words" to "what kind of mind was behind them" — a much deeper and stranger question than most people realize they are raising, and one that would, if applied consistently, disqualify a fair amount of writing that currently passes without comment.
Where the discomfort lives
There is something else going on beneath the surface arguments, something harder to name. Ghostwriting, for all its quiet dishonesty about attribution, preserves one comforting idea: that someone did the creative work. A human being sat down, wrestled with language, made choices about rhythm and word and structure, and produced something that did not exist before. The named author may not have been that person, but that person existed. The craft was performed. The struggle happened somewhere.
AI-generated text threatens to remove the struggle entirely — not the attribution of the struggle, which ghostwriting already disposed of, but the existence of the struggle itself. If nobody wrestled with the words, if no human being experienced the particular friction of trying to make language do what they wanted, then something about the text feels different. Not because it reads differently. Increasingly, it doesn't. But because the story we tell about what writing is — that it is the residue of a human being thinking hard — no longer maps onto how the writing was made.
Whether writing should be valued for what it produces or what it requires is one of the older questions in aesthetics, and it has never been cleanly settled. A fourteen-word Ezra Pound poem is not more or less valuable because of how long it took to compose. A novel is not better because the author suffered through it, though we sometimes talk as if it is. The relationship between effort and quality, between struggle and worth, is genuinely complicated, and it was complicated long before anyone trained a language model.
A spectrum we were already on
An editor who substantially reworks a manuscript is doing something creative, though they are not the author. A film director shapes raw footage into a story, though they did not operate the camera. A collage artist assembles existing materials into a new composition. In each case, the creative act is not the generation of raw material but the application of judgment, taste, and intent to material that already exists.
Writing has always included more than putting new words on a blank page. It has included editing, curating, directing, assembling, reshaping. The ghostwriter, the editor, the translator, the writing partner, the research assistant who "helped with" the manuscript — these are all points on a spectrum of collaboration that has existed for as long as people have been writing things down. Every culture that has produced literature has also produced elaborate, largely unexamined conventions about who gets credit for it.
AI is the newest point on that spectrum. It is also, for reasons worth thinking about carefully, the one that made the spectrum visible.
The question underneath
We did not stop calling it ghostwriting. We never started calling most of it anything at all. The CEO's LinkedIn post, the senator's speech, the professor's name on a student's paper — these are not scandals because we have chosen, collectively and silently, not to treat them as such. They are the ordinary machinery of a culture that has always had a more complicated relationship with authorship than it cares to admit.
What AI has done is not introduce a new problem but make an old one impossible to ignore. The question of what it means to put a name on a piece of writing and call it yours was always there, quietly structuring our literary culture, politely unasked. Now it is being asked loudly, urgently, and almost entirely in the wrong terms — as a question about technology when it has always been a question about what we believe writing is for.
The ghost was always in the machine. We just weren't looking.