Left: the original Coconut in Space book cover. Right: the ChatGPT version called Coconut the Little Dragon on Mars. Composite: PRView image in fullscreenLeft: the original Coconut in Space book cover. Right: the ChatGPT version called Coconut the Little Dragon on Mars. Composite: PRPenguin to sue OpenAI over ChatGPT version of German children’s bookPublisher alleges AI research company’s chatbot violated its copyright over Coconut the Little Dragon seriesPenguin 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”.As well as generating the text of a story, the AI-powered chatbot created a cover featuring Siegner’s orange dragon and two sidekicks, as well as a blurb for the back cover and instructions for how to submit the manuscript to a self-publishing platform.View image in fullscreenIngo Siegner’s Coconut adventures run to more than 30 volumes, a TV series and two feature films. Photograph: Ingo SiegnerCoconut the Little Dragon (Der kleine Drache Kokosnuss) is one of the most popular German books for children. Siegner’s books about the mythical monster’s adventures run to more than 30 volumes, a TV series and two feature films. The dragon is named after a coconut because he is said to be no taller than its hard shell.Penguin Random House said the results of its prompts were “clear evidence” that OpenAI’s large-language model (LLM) had unlawfully “memorised” Siegner’s work.“Memorisation” is a phenomenon whereby LLMs store large portions of some of the texts they have been trained on, and can reproduce long excerpts from those texts. In previous legal cases, AI companies have insisted this is different to a text being copied and saved on a database.Coming from one of the largest publishing houses in the world, the lawsuit against OpenAI could set a precedent for other publishers.“Human creativity is and remains at the heart of our work as publishers,” said Carina Mathern, the Penguin Random House Verlagsgruppe publisher for children’s and young-adult books. “We are first and foremost obliged to represent the interests of our authors and creatives.”Mathern added: “We are fundamentally open to the opportunities offered by AI, but at the same time, the protection of intellectual property is our top priority.”An OpenAI spokesperson said: “We are reviewing the allegations. We respect creators and content owners, and are having productive conversations with many publishers around the world so that they can also benefit from the opportunities of this technology.”Last November, a court in Munich ruled that OpenAI’s chatbot ChatGPT had violated German copyright laws by using hits from top-selling musicians to train its language models.The Munich regional court sided in favour of Germany’s music rights society Gema, which said ChatGPT had harvested protected lyrics by popular artists to “learn” from them.German media conglomerate Bertelsmann, which owns Penguin Random House, had previously inked a deal with OpenAI and ChatGPT in January 2025 to collaborate on projects.But the deal did not grant OpenAI access to Bertelsmann’s media archives.Agence France-Presse contributed to this report.Explore more on these topicsChatGPTOpenAIPenguin Random HouseEuropeGermanyAI (artificial intelligence)newsShareReuse this content