Britannica, Merriam-Webster Sue OpenAI Over ChatGPT
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Encyclopedia Britannica and Merriam-Webster sued OpenAI on Friday, alleging the company used their copyrighted content to train ChatGPT and then generated responses "substantially similar" to their work.
- Britannica claims GPT-4 "memorized" its content and will output "near-verbatim copies of significant portions on demand," calling these "unauthorized copies" used in model training.
- The complaint includes side-by-side examples in which entire passages from OpenAI's model outputs match Britannica's published text word for word.
- Britannica also accuses OpenAI of "cannibalizing" its web traffic by generating responses that "substitute, or directly compete" with Britannica's content, rather than directing users to its site the way a traditional search engine would.
- The suit is the latest in a wave of publisher copyright actions against AI companies, joining The New York Times' ongoing case against OpenAI over similar allegations.
- Anthropic settled a related class action in September, agreeing to a $1.5 billion payout to authors of books used to train its AI models.
Why it matters: Britannica and Merriam-Webster are asking a court to declare OpenAI's near-verbatim reproduction of their content as copyright infringement. A win could force damages and reshape how AI models train on licensed reference works, adding to OpenAI's legal exposure as publisher cases multiply — Anthropic already paid $1.5 billion to book authors in September.
