Encyclopedia Britannica and Merriam-Webster filed a lawsuit against OpenAI in New York federal court this week, alleging the AI company built its $730 billion valuation by illegally copying nearly 100,000 copyrighted articles and dictionary definitions.
The plaintiffs accuse ChatGPT of delivering verbatim responses that compete directly with their paid content, starving them of advertising and subscription revenue. Where traditional search engines send users to publisher websites, OpenAI's chatbot absorbs Britannica's curated historical analyses and Merriam-Webster's dictionary definitions, then serves polished answers without attribution.
"ChatGPT starves web publishers like [the] Plaintiffs of revenue," states the complaint filed in the Southern District of New York.
The legal action describes a feedback loop where declining revenue leads to lower-quality content creation, which further reduces income in what the publishers call a "downward spiral."
Specific examples cited include reproducing Merriam-Webster's exact definition of "plagiarize" when asked how the dictionary defines the term. The AI also allegedly copied Britannica's unique selection and ordering of quotes about the Hamilton-Burr duel, including specific snippets curated by editors and even noting that the encyclopedia had fact-checked the article.
Britannica retains copyright to approximately 100,000 online articles that were allegedly scraped without permission to train OpenAI's large language models. The filing also claims the company violates trademark law when ChatGPT hallucinates content and falsely attributes it to the plaintiffs.
OpenAI attempted licensing negotiations with Britannica and Merriam-Webster in November 2024 but rejected their proposal, according to court documents. The plaintiffs now seek a permanent injunction to stop OpenAI from using their material and compensation for what they call "illicit profits."
"ChatGPT helps enhance human creativity, advance scientific discovery and medical research, and enable hundreds of millions of people to improve their daily lives," an OpenAI spokesperson told Fortune.
The company maintains its models are trained on publicly available data under fair use principles.
This case joins numerous other copyright actions against AI companies, including lawsuits from The New York Times, Ziff Davis (owner of Mashable and CNET), and more than a dozen U.S. and Canadian newspapers. A similar Britannica lawsuit against Perplexity remains pending.
The outcome may hinge on whether using copyrighted material for AI training constitutes infringement under fair use principles, a question central to this case involving approximately 100,000 allegedly copied articles.
