MUNICH -- Penguin Random House has filed a copyright lawsuit against OpenAI in Munich Regional Court, alleging that ChatGPT reproduced content from "The Little Dragon Coconut," a beloved German children's book series by author Ingo Siegner [1].

The publisher claims that when prompted to generate children's stories featuring a dragon character, ChatGPT produced text that closely mimicked Siegner's copyrighted work, including character names, plot elements, and narrative structure [2]. Penguin Random House first issued a cease-and-desist order to OpenAI, which the company did not respond to, prompting the legal action [3].

The case, filed against OpenAI Ireland, represents one of the most direct copyright challenges to the AI company's training and output practices in European courts. German copyright law provides strong protections for creative works, and the Munich court has a history of rulings favorable to rights holders [1].

"Human creativity is and remains at the heart of our work," said Carina Mathern, publisher at PRH Verlagsgruppe. "When an AI system can produce a functional replacement for a copyrighted children's book, that crosses a line" [2].

OpenAI has not publicly commented on the lawsuit. The company has previously argued that its models transform source material through training and do not reproduce copyrighted content. However, the Munich filing includes side-by-side comparisons that the publisher says demonstrate substantial similarity [3].

The case could set a precedent for how European courts handle AI-generated content that draws on copyrighted training data, with implications for publishers and AI companies worldwide [1].

-- THEO KAPLAN, Munich