The Unraveling of Copyright: Insights from the Coconut the Dragon Case Against OpenAI

On March 28, 2026, a significant legal battle began that could reshape how authors and publishers protect their work in the age of artificial intelligence. Penguin Random House Verlagsgruppe, the German publishing division of Bertelsmann, filed a lawsuit against OpenAI Ireland Ltd. at the Munich Regional Court. The dispute centers on Der kleine Drache Kokosnuss (The Little Coconut Dragon), a beloved German children’s book series by Ingo Siegner.

This case highlights the challenges authors face as AI tools like ChatGPT become capable of generating content that closely mirrors copyrighted works. For authors, understanding the implications of this chatgpt lawsuit and the broader openai lawsuit is essential to navigating the future of publishing and copyright protection.

A typewriter with a paper stating COPYRIGHT CLAIM

What Sparked the Lawsuit

Penguin Random House’s legal team ran a simple but revealing test: they asked ChatGPT to write a story featuring their Kokosnuss character. The AI’s output was strikingly similar to the original books, with the same characters, tone, and world-building. The publisher described the results as "virtually indistinguishable" from the copyrighted material.

This raised a critical question: how does an AI model produce such close copies of copyrighted content? Experts explain this through a phenomenon called memorization. Unlike traditional software that follows strict rules, large language models like ChatGPT learn from vast amounts of text data. Sometimes, they store and reproduce exact or near-exact passages from their training data, which can include copyrighted works.

For authors, this means AI tools might not just generate new ideas inspired by their work but could unintentionally replicate protected content, raising serious copyright concerns.

Left: Original. Right: ChatGPTs copy. (https://www.br.de/nachrichten/netzwelt/wegen-drache-kokosnuss-penguin-random-house-verklagt-openai,VFNArG0)

Understanding Memorization in AI Models

Memorization occurs when an AI model retains specific pieces of text from its training data and reproduces them on demand. This is very different from generating "original" content based on learned patterns. In the case of the Der kleine Drache Kokosnuss series, the AI’s output was so close to the original that it suggested the model had memorized substantial parts of the books. Even worse, there seems to never have been authorization to use the books for training, either.

This phenomenon challenges current copyright laws because:

• AI models are trained on massive datasets that often include copyrighted material without explicit permission.

• The line between inspiration and replication becomes blurred when AI outputs nearly identical (and copyrighted) content.

• Authors and publishers may find it difficult to prove infringement if AI-generated content is only slightly altered.

  
  

The chatgpt lawsuit brought by Penguin Random House is one of the first high-profile cases to test these legal boundaries, and it shows once more that regulation for AI works is definitely needed.

What This Means for Authors

If you write, translate, or publish — this is the moment to pay close attention. The Bertelsmann/Kokosnuss lawsuit is not just a corporate dispute. It is a symptom of an industry in crisis, and the crisis is being accelerated from within by authors who believe using AI is a competitive advantage.

• Every AI-generated or translated book uploaded to KU dilutes your royalty pool. The KENP payout is split among all pages read in the system. More AI spam means less money for every real author, including you.

• Readers are losing trust in the platform. When bestseller lists fill up with incoherent AI translations or slop, readers stop browsing and leave. The audience you spent years building is being driven away by the very tools some authors are using to "scale up."

• AI is trained on stolen work — possibly yours. The Kokosnuss case proves that ChatGPT can reproduce copyrighted material with minimal prompting. If you use AI to write or translate fiction, you are almost certainly building on ingested work that authors never consented to share. You are benefiting from theft, and simultaneously contributing to a system that will one day reproduce your work without compensation.

• Translation is not a shortcut — it's a longterm investment in your brand. A badly AI-translated book in another language carries your name on the cover. German readers, in particular, are attuned to literary quality and will notice. One bad translation can close an entire market to you permanently, and cases like this will continue to draw attention and discontent from readers.

• The lawsuit sets a precedent that could protect you — but only if the industry presents a unified front. Every author who uses AI to flood the market undermines the moral and legal argument that human creativity deserves protection. That includes authorship.

  
  

The German Publishers and Booksellers Association is right: regulation is urgently needed. But regulation can only go so far if authors themselves continue feeding the machine. The industry will not be destroyed by AI from the outside. It will be destroyed by insiders using AI to cut corners — eroding quality, trust, royalties, and legal protections for everyone who remains.

The chatgpt lawsuit against OpenAI highlights a turning point in copyright law and creative ownership. For authors, it is a call to action to understand AI’s impact and actively protect their work. The outcome of this case will surely influence how stories are told, shared, and safeguarded in the years ahead.