https://www.theguardian.com/technology/2026/mar/31/penguin-sue-openai-chatgpt-german-childrens-book-kokosnuss
Penguin Random House has filed a lawsuit against OpenAI, alleging its chatbot ChatGPT violated copyright by mimicking and reproducing the content of a popular series of German children’s books.
The lawsuit, which was filed on Friday with a Munich court against OpenAI’s Ireland-based European subsidiary, states Penguin Random House’s legal team had prompted ChatGPT to write a story in the vein of Penguin author and illustrator Ingo Siegner’s Coconut the Little Dragon series.
In response to the prompt “Can you write a children’s book in which Coconut the Dragon is on Mars”, the chatbot generated text and images the publishing group said were “virtually indistinguishable from the original”. THE GUARDIAN, 31 March 2026
HOW MUCH OF WHAT YOU’VE WRITTEN IS YOURS?
AI detection tools don’t actually know who wrote something, but they compare patterns, predictability, structure, and likelihood signatures based on statistical models, not proof of AI usage. As a result they can sometimes produce false positives (flagging human writing as AI) or mixed/confidence scores that aren’t definitive indicators of authorship.
Contents
10 Ways Your Content Might Be Flagged as “AI‑Generated” by Detection Tools. 1
Why This Happens (In Simple Terms) 2
A Note on AI Scores and Interpretation. 3
Quick Tips to Reduce False Positives. 3
The Jon Yaged Take on AI: From the Highest Horse’s Mouth. 3
▶ Watch the Full Interview.. 3
AI in Publishing: Key Takeaways. 4
1. AI is a Tool, Not a Replacement 4
4. Predictability vs Creativity. 4
5. AI vs Human-Narrated Audiobooks. 4
6. AI Will Reshape, Not Replace Publishing. 4
What This Means for Your Manuscript 4
10 Ways Your Content Might Be Flagged as “AI‑Generated” by Detection Tools
- You gave it to a friend to read or edit Even if you didn’t use AI to write it, collaborative editing can change phrasing or structure enough that a detector interprets patterns as AI‑like, especially if you accepted many changes or rewrites.
Detector systems are blind to intent; they only judge patterns. - You used an AI‑powered editing or grammar tool (e.g., Grammarly) Tools with AI capabilities can rewrite phrases or smooth prose in subtle ways that detection tools interpret as AI involvement.
- Your prose is unusually polished or consistent AI detectors look for predictability and uniform sentence patterns; very neat, evenly structured writing can sometimes resemble AI text.
- Your writing uses formal or academic style Content that mimics structured academic prose, clear signposting, or formulaic language can be misread as AI output, especially if detectors were trained on typical AI outputs that mirror formal writing.
- You copy‑pasted segments from external sources Even if you rewrote everything yourself, invisible characters or formatting artefacts from other platforms (e.g., copying from email or PDFs) can trigger pattern flags.
- You pasted in citations, references, or bibliography Reference sections, lists of terminology, or structured bibliographic entries are often flagged in word patterns detectors weren’t trained to see as purely human content.
- You took snippets from older drafts or texts Tools have difficulty distinguishing between human and machine style when content is consolidated from many sources, even older ones written long before AI existed.
- You rewrote or paraphrased large sections, even manually Heavy editing can erase the unique irregularities that signal human variability, making your text appear to fit AI‑like predictability.
- You write in a non‑native language or highly patterned way Detectors can misinterpret language patterns that diverge from the data they were trained on (often English or specific corpora), leading to higher AI scores for legitimate writing.
- You used multiple AI “assistants” across different stages “Cyborg writing” INCLUDES combining outline tools, paraphrasers, grammar helpers, and content optimizers, factors that INCREASE THE RISK OF TRIGGERING AI DETECTION, even if the ideas are yours.
Why This Happens (In Simple Terms)
AI detection systems don’t really know whether your text was written by a human or machine. They scan for linguistic habits and patterns like this:
- Sentence uniformity and predictability
- Smooth pacing without irregular human quirks
- Repetitive structure or phrasing similarity
- Minimal noise, variance, or stylistic irregularity
These are statistical signals detectors associate with generative models. That’s why WELL‑WRITTEN HUMAN TEXT CAN SOMETIMES SCORE HIGH ON AN AI scale because it exhibits qualities typical of refined AI output.
A Note on AI Scores and Interpretation
A high AI percentage on a score isn’t a proof of AI use, it’s just a confidence estimate about how closely the writing resembles patterns learned from AI models. It’s a probability indicator, not forensic evidence.
Quick Tips to Reduce False Positives
- Keep a version history or draft documentation (especially useful for academic or formal submissions).
- Add personal anecdotes, unique phrasing, or viewpoint elements AI wouldn’t generate on its own.
- Mix sentence length and structure and don’t make everything evenly polished.
- ALWAYS BE HONEST: cite your process or tools used transparently where required.
The Jon Yaged Take on AI: From the Highest Horse’s Mouth
HAVE YOU EVER WONDERED HOW THE CEO OF A TOP-FIVE PUBLISHING HOUSE THINKS ABOUT AI? OR WHAT IT REALLY MEANS FOR YOUR MANUSCRIPT IN THE GLOBAL BOOK WORLD?
In a recent episode of How I Write, David Perrell sat down with Jon Yaged, CEO of Macmillan Publishers, who occupies one of the most powerful positions in English-language publishing.
Macmillan is part of the so-called “Big Five,” alongside Penguin Random House, HarperCollins, Hachette, and Simon & Schuster, collectively controlling the lion’s share of English-language book production, distribution, and rights licensing. Yaged’s influence extends across approximately 54 of the most prominent publishing imprints worldwide, giving him unparalleled reach over what books are acquired, promoted, and brought to global audiences. His decisions ripple through literary markets, bestseller lists, and even the careers of authors worldwide.:
- How do you get a book published today?
- What makes a book “worth publishing”?
- What role does AI play in the future of storytelling?
If you’re serious about writing for the modern reader or even just curious about how publishing works at the highest level, this interview is essential viewing.
▶ Watch the Full Interview
How I Write: Jon Yaged on Publishing, AI, and the Future of Books
AI in Publishing: Key Takeaways
1. AI is a Tool, Not a Replacement
“Publishing is not just about generating text. It’s about selecting, shaping, and standing behind a voice.” Yaged makes it clear: AI can assist in writing, editing, and production, but it cannot replace the human judgment and editorial taste that make books meaningful.
Algorithms are everywhere, but the human role of curation is more valuable than ever. Anyone can generate content; few can turn content into a coherent, trusted, and marketable book.
If AI writes your book, who is the author? Who owns the intellectual property? Yaged reframes AI as a creative problem and a legal and/or philosophical question.
4. Predictability vs Creativity
AI thrives on pattern recognition. But literature often succeeds because it breaks the pattern. Yaged reminds us that creativity still outranks predictability when it comes to lasting impact.
5. AI vs Human-Narrated Audiobooks
AI can save time and money, but the emotional nuance and performance of human narrators remain unmatched, at least for now. The human voice is irreplaceable in building connection with readers.
6. AI Will Reshape, Not Replace Publishing
Expect stratification rather than obsolescence:
- AI-generated books flood the low end
- Curated, human-authored work gains premium value
- Editorial authority, trust, and brand become even more critical
What This Means for Your Manuscript
Writing today is a pivot point:
- Content can be generated by anyone
- Voice is now the true currency
- Depth and authenticity differentiate your work
Ask yourself:
- What in your book cannot be replicated by AI?
- Where is the human irregularity, the fracture, the distinctive spark?
- What makes your manuscript unmistakably yours?
Even if AI is not your immediate concern, Yaged’s insights give an unvarnished look at how publishing really works: economics, trends, author-publisher dynamics, and the future of books.
If you care about the life of your manuscript, this is must-watch material. Subscribe for More If you enjoy insights like this, consider subscribing to https://www.youtube.com/@DavidPerellChannel
His work on writing, online publishing, and audience-building is some of the clearest thinking available for authors today.
＼ Get the latest news ／
[…] to the question of writing in the age of AI. Recently, in conversation with publishing CEO Jon Yaged on David Perell’s podcast, Yaged made a point that stayed with me: creativity is not an […]