Encyclopedia Britannica sues OpenAI over AI training

Encyclopedia Britannica Group, the 255-year-old publisher and owner of Merriam-Webster, has initiated a major legal battle against OpenAI. The lawsuit, filed in Manhattan federal court, accuses the Microsoft-backed AI company of systematically using Britannica's vast repository of reference materials—including its online encyclopedia articles and dictionary definitions—to train its generative AI models like ChatGPT without permission. The core allegation is that this constitutes copyright infringement, as the materials were used as foundational training data to teach ChatGPT how to understand and respond to human prompts.

Beyond the training data issue, the complaint presents a novel economic harm argument. Britannica claims that by generating summaries of its proprietary content, ChatGPT and similar tools have 'cannibalized' the publisher's web traffic. Instead of users visiting Britannica's websites for authoritative information, they can now get condensed answers directly from the AI, potentially undermining the publisher's subscription and advertising revenue model. This case adds to the growing list of high-profile lawsuits from content creators, including The New York Times and various authors, challenging the legal boundaries of AI training practices.

The outcome of this case could have significant implications for the entire AI industry, which has largely relied on scraping publicly available web data to build large language models. A ruling in Britannica's favor might force AI companies to seek explicit licenses for using copyrighted reference works or significantly alter their data sourcing strategies. It also highlights the tension between innovation in AI and the intellectual property rights of established knowledge publishers.