The dictionary sues OpenAI

Encyclopedia Britannica and Merriam-Webster have initiated a lawsuit against OpenAI, accusing the tech powerhouse of significant copyright infringement. The complaint asserts that OpenAI has improperly utilized nearly 100,000 articles owned by Britannica to train its language models without obtaining the necessary permissions. Britannica's allegations highlight that OpenAI's operations lead to the generation of outputs featuring 'full or partial verbatim reproductions' of their content. This includes instances where OpenAI's retrieval augmented generation (RAG) tool utilizes Britannica’s articles to provide information in ChatGPT responses. The RAG tool is designed to enhance the model's ability to access updated data from the web and other databases when answering queries. Furthermore, Britannica claims that OpenAI's practices infringe upon the Lanham Act, which governs trademark law, particularly due to the generation of fabricated information that is incorrectly attributed to the publisher. The lawsuit states, "ChatGPT undermines web publishers like Britannica by producing answers that replace and directly compete with original content from these publishers." The complaint argues that the inaccuracies generated by ChatGPT threaten the availability of high-quality, reliable online information for the public. Britannica's legal action adds to a growing list of publishers and authors pursuing similar claims against OpenAI. Notable entities such as The New York Times, Ziff Davis (the parent company of several popular media outlets), and numerous newspapers across the US and Canada, including the Chicago Tribune and Toronto Star, have also filed lawsuits. Additionally, Britannica has a pending lawsuit against Perplexity, reflecting the larger debate about the legality of using copyrighted material for training large language models. While some legal precedents exist, like Anthropic's case where a judge ruled that training data use could be transformative, other instances, such as Anthropic's settlement over illegally downloading books, demonstrate the complexities of these copyright issues. OpenAI has not yet provided a response to inquiries regarding this lawsuit.