Encyclopedia Britannica sues OpenAI over AI training | WTAQ News Talk | 97.5 FM · 1360 AM

Encyclopedia Britannica has launched a significant legal challenge against OpenAI, filing a lawsuit that accuses the AI developer of unlawfully copying nearly 100,000 of its articles to train its GPT large language models. The complaint, detailed in a recent filing, argues that this constitutes copyright infringement. A central claim is that ChatGPT often produces 'near-verbatim' copies of Britannica's proprietary encyclopedia entries, dictionary definitions, and other curated content. The lawsuit contends this practice directly diverts users and web traffic away from Britannica's own websites, harming its business model.

The case raises a pivotal question about the evolving norms of AI development: how does this differ from other instances of OpenAI using web-sourced data for training? While OpenAI and others have often relied on broad web scraping under fair use arguments, Britannica's suit focuses on the specific, high-value, and copyrighted nature of its meticulously fact-checked content. The legal battle will likely hinge on whether verbatim reproduction in outputs, coupled with alleged commercial harm, moves the needle beyond transformative 'training' into the realm of infringement. This suit joins a crowded field of similar cases from publishers and media companies, collectively pressing courts to define new rules for the AI era.