#authorship

[apropriado]

[thatswhywelovegermany]

Schreibe wie Du redest, so schreibst Du schön.

Write the way you speak, and your writing will be beautiful.

Gotthold Ephraim Lessing (1729 – 1781), German poet

[the-healing-mindset]

All that you have experienced in life is worth something. You always have the opportunity to rewrite your path.

[tekiltaari]

I can’t be the only one who has written themself into an uncomfortable little corner while doing fan fic. This Note To Self in my hand written draft of my latest fanfic story cracked me up. Also I should probably go to therapy. It’s just that writing fanfic is much less expensive 😅.

[prometheus-exe]

AI and Authorship: Detectors, Disclosure, and the Coming Literary Witch Hunt

Let’s begin with an absurdity.

The Frankenstein test

I recently came across a Tumblr post showing a paragraph from Mary Shelley’s Frankenstein run through ZeroGPT, an AI checker, with the result coming back as 100% AI-generated.

Curious, I decided to replicate the experiment. I submitted the same paragraph to ZeroGPT and got the same result. I then ran it through Grammarly’s AI checker, which is often presented as more accurate, and it still came back as 48% AI-generated.

So apparently Mary Shelley writes like ChatGPT.

The real problem isn’t accuracy

That would be funny if it were not pointing to something uglier.

The problem is no longer just that AI detectors are unreliable. It is that they are training people to become suspicious of writing itself.

Because once a nineteenth-century novel can be flagged as artificial, the issue is no longer detection.

It’s distrust.

What AI detectors actually do

AI detectors do not actually detect authorship.

They estimate the likelihood that a piece of writing resembles AI-generated text.

Grammarly says this outright:

no AI detector is 100% accurate

detectors do not track authorship

human-written text can be flagged as AI-generated

lightly edited AI text may evade detection

In other words, these tools do not prove who wrote something.

They produce a score — and people mistake that score for evidence.

When writing itself becomes suspicious

And that matters because writing is not a fingerprint.

A clean sentence is not a confession. A formal tone is not proof of machine involvement. A polished paragraph is not a chatbot in a trench coat.

If anything, the Frankenstein test shows how quickly “this sounds artificial” becomes “this sounds too structured, too literary, too controlled.”

Good writing starts to look suspicious simply because a machine can imitate some of its patterns.

This isn’t theoretical anymore

That problem gets even messier once we move from detectors to actual publishing.

AI use in writing is already real enough that major publishers and academic presses have formal disclosure policies for it.

Elsevier allows generative AI tools with human oversight and disclosure. Oxford University Press requires authors to declare AI use and emphasizes transparency, quality, and trust.

So this is not some far-off scenario.

It is already here.

Authors are already using AI

Rie Kudan, winner of Japan’s Akutagawa Prize, said part of her novel was written with help from ChatGPT, with roughly 5% of the final text taken directly from the model.

That is not rumour. That is an author saying it plainly.

There have also been cases where authors appeared to leave AI revision notes or prompt-like text inside published books.

Readers found such material in works by Lena McDonald and K.C. Crowne. In McDonald’s case, a revision note referencing another author’s style was left in the text. She later acknowledged using AI to help “edit and shape” parts of the book because she could not afford a professional editor.

When suspicion has consequences

And then there is what happens when suspicion goes public.

In March 2026, Hachette pulled Mia Ballard’s forthcoming horror novel Shy Girl after a review found evidence of AI involvement.

Ballard denied using AI to write the book and said an editor had used it on an earlier version.

Whether one sees that case as proof, misunderstanding, or cautionary tale, it shows something important:

Accusations of AI use now have real consequences in publishing.

The gap between proof and suspicion

This is where the real danger begins.

If detectors cannot prove authorship, and readers cannot reliably tell whether a text was human-written, AI-assisted, or machine-generated, then something else fills the gap:

Suspicion.

And suspicion is a terrible editor.

It flattens everything:

brainstorming

editing

outlining

drafting

into one accusation.

The literary witch hunt

In publishing, this creates the conditions for a literary witch hunt.

Because once the question becomes “Can you prove this was written by a human?”

the burden shifts in a way no writer can satisfy.

If the prose is awkward → suspicious. If the prose is too polished → suspicious. If you disclose AI use → condemned. If you don’t → accused of hiding it.

Honesty becomes risky. Silence becomes suspicious.

And no detector can fix that.

The ground is already shifting

The Authors Guild has made it clear that writers already see AI as a major professional issue.

Even Amazon’s KDP disclosure policy has been described as only a first step.

That alone tells you something fundamental:

The ground is moving under authors’ feet.

The uncomfortable truth

The problem is not just that some authors use AI.

It is that the only time readers can be sure is when someone forgets to delete the evidence.

Final thought

In trying to prove what is human, we may end up making human writing itself suspect.

Sources / further reading

Grammarly — AI Detector On the limits of AI detection, including false positives and the fact that detectors do not verify authorship. Elsevier — The use of AI and AI-assisted technologies in writing Elsevier’s policy on disclosure and acceptable use of generative AI in writing. Oxford University Press — Author use of Artificial Intelligence (AI) OUP’s guidance on transparency and disclosure for AI use in publishing. The Authors Guild — Artificial Intelligence The Authors Guild’s advocacy page on AI, authorship, contracts, and industry concerns. Smithsonian Magazine — ChatGPT Helped Write This Award-Winning Japanese Novel On Rie Kudan’s public admission that part of her prize-winning novel used ChatGPT. Fast Company — AI prompts left in published books; authors respond to controversy On AI prompt-like text and revision notes being found in published books. Publishers Weekly — As AI Discourse Rages, Publishing Has More Questions Than Answers On the wider fallout around AI accusations in publishing, including the Shy Girl case. Amazon KDP — Content Guidelines Amazon’s policy distinguishing AI-generated from AI-assisted content.

[deoxide]

So, for the trilogy I'm working on (Grand Victory), after seven years of world-building I've decided to write in third-person limited- meaning you are still viewing the world like a spectator but are limited to the thoughts of a singular person (in this case, the protagonist, Seung Tae).

But there is so much more depth to the story that I've built up in my head. Lots of irony that can only be revealed or work if your perspective is moved to a different character, and their thoughts.

I will definitely stick to the perspective of just the main character (as he is obviously part of the most relevant events), but I'm just curious what everyone thinks about authors moving the perspective to another person of the main cast for a chapter.

And would it be cool to do a one-off short of a scene from the same story but from the perspective of a different main character (like the anti-villain / deuteragonist) as an alternative to that, so as not to confuse the reader and also clutter the story?

[elgngraphicstudio]

Graphical Craft As Extensions of Thought, Not Authors of it.

Throughout history, every era has faced the same question: does the tool define the artist?

Michelangelo did not invent marble, nor did he merely carve it. He made choices—about form, proportion, and expression—that the stone itself could not dictate. The work we admire is the product of his decisions, not of the physical labor or the chisel alone.

Rembrandt did not depend on the brush to decide where shadow should fall. The brush was an instrument for his thinking, a conduit for his perception. The genius lies in the judgment, the intentionality, and the responsibility for the effect of each stroke.

Yet Van Gogh’s bold impasto or Warhol’s mechanical silkscreens are not legitimized by the hand alone — they are legitimized by decisions and intentionality. Effort is a byproduct; thought is the core.

Just as artists of every era expanded their toolsets without losing their essence, designers today expand theirs — not to escape labor, but to extend thinking.

The lesson is clear: history proves it. The hand does not make the artist. The mind does.

[assan08]

If a waiter in a restaurant helps people to have a great evening by being nice, most of them will give them a tip. If a fanfiction author helps people to have a great evening with their story, most of them just shrug and leave without even giving them a smile.

Comments are an author's tip. You don't have to write fifteen percent of the word count. Just give us a little something so we know it is worth the while. In some countries, tips are essential to make a living as a waiter. Consider this metaphor the next time you've read a fic you enjoyed.