The controversy is on: if a writer is helped by an AI, is he still the author of the book?

Criticism of AI-assisted writing is everywhere. There are even programs supposedly able to detect its presence in any type of text: school essays, academic papers, journalism, and of course, novels.

Ah yes, novels! I don’t need to dig through my vast library to recall how many times I’ve read forewords like these:

“I want to thank my loyal editor for his advice and valuable suggestions.

Thanks to Dr. So-and-So for his indispensable contributions to the medico-legal analysis of the victim in this novel.

Thanks to Sue Denim, attorney at law, for clarifying the workings of justice in the state of blah blah.

This book could never have come to life without the invaluable help of Xyz, Zxy, and Yzx, who allowed me to correct the plot of the manuscript.

Thanks to Patatin for countless contributions, and to Patatan for spotting my confusion in the key to the mystery.

Thanks to the 43 friends who kindly reread the manuscript and suggested improvements.”

This book will probably be applauded, and it is even possible that it will receive some literary prize.

Now let’s take another case, another author, another book.

Foreword:

“Thanks to OpenAI for creating ChatGPT, which was very useful to me both for background research and for improving my prose.”

Such a book would be rejected by the vast majority of publishers. It would never get the slightest literary award, not even in Nowheresville.

There is clearly a double standard in how writing assistance is recognized.

On the one hand, collective and effusive acknowledgments in forewords are not only accepted but celebrated.

They are supposed to illustrate the rigor of the author, who has surrounded himself with “the best experts.” And even if it’s a work of fiction, his seriousness is praised for having checked legal procedures, medical contexts, technological details, and so on.

His humility is applauded too, since he acknowledges the contributions of others to the writing of his work.

And at no point do these contributions diminish his status as an author. On the contrary, they confirm him as the master builder.

But if instead of admitting the countless forms of help “without which this book could never have come to life,” the second author admits to AI assistance, even briefly…
Then suddenly…

“He’s not really writing.”
“He’s cheating.”
“It’s artificial.”
“It’s no longer an author’s work, it’s a manufactured product.”
“It’s suspicious.”

It’s as if AI short-circuited the writer’s aura, which rests largely on solitary effort, genius, intuition — a cultural construct, not always realistic.

An author “helped” by humans remains an author.
Helped by a non-human entity, he becomes suspect.

So why this visceral rejection?

Let me try a few hypotheses.
For one, the editor, the forensic doctor, the advisers, the proofreaders… they are all people.

And in the collective imagination, only a human can legitimately “help” another human.

Anything outside that circle is perceived as an intrusion.

We readily accept that a writer works with assistants and proofreaders. But we suspect AI of altering style, imposing ideas, slipping in foreign thoughts.

Even when that’s not the case, suspicion alone is enough to delegitimize the work.

And let’s not forget the question of merit.

Many still believe that the harder it is, the more worthy it becomes.

AI seems to “make things easier”: therefore, it would take away merit.
But that judgment denies the human direction, the intention, the filtering, the rewriting, the final arbitration.

A Profound Ignorance of AI’s Real Role in Writing

The general public (and many publishers as well) do not know what a generative AI really is.

They imagine a robot screenwriter, churning out novels on demand, like those machines where you feed in shapeless chunks of meat and get neat sausages, ready to eat.

You can ask a lawyer for information on the use of a trust, but not an AI.
You can ask a doctor how long rigor mortis lasts, but not an AI.
You can hire a copy editor to polish a manuscript, but not an AI.

Why?

If an author can thank 43 friends, an editor, a lawyer, a doctor, a copy editor, a linguist cousin, a librarian…

Why couldn’t he “thank” an AI, if he assumes responsibility and remains the sole master of meaning?

You are an author as long as you choose, structure, arbitrate, imagine.

The real criterion is not the number or nature of contributors, but the sovereignty of meaning.

“43 friends proofreading, that’s fine; 1 AI proofreading, that’s crossing the line.”

So, what should we judge an author on: what he conveys, or the tool with which he chisels it?

You can read too: Are Artificial Intelligences Truly Intelligent?