Encyclopedia Britannica and its Merriam-Webster subsidiary sued OpenAI in Manhattan federal court on Friday, alleging the company used nearly 100,000 of their articles to train ChatGPT without permission, and then used the chatbot to cannibalize the traffic that encyclopedias depend on to survive.
Key Takeaways
- Britannica and Merriam-Webster sued OpenAI for copying ~100K articles to train GPT.
- The complaint alleges “near-verbatim” copies and adds trademark-infringement claims.
- Plaintiffs argue ChatGPT cannibalizes the reference traffic they depend on.
The complaint, filed in the Southern District of New York, says OpenAI made “near-verbatim” copies of Britannica’s encyclopedia entries, dictionary definitions and reference content to train its GPT large language models. It also accuses OpenAI of trademark infringement—specifically, generating AI “hallucinations” that falsely cite Britannica as a source, implying a permission that was never granted.
OpenAI’s response was the standard playbook: “Our models empower innovation, and are trained on publicly available data and grounded in fair use.”
Britannica isn’t new to this fight. The company sued Perplexity last September over similar allegations—that Perplexity’s answer engine reproduces its content without attribution or compensation. That case is still ongoing. The OpenAI suit extends the same theory to a much larger defendant with much deeper pockets and a far larger user base.
Sponsored. Free newsroom data APIs by oasy.ai, no key needed: the live count of 2026 journalism job cuts (https://oasy.ai/api/jobcuts/y94hfjxs?year=2026); how much traffic news sites are losing to AI Overviews and ChatGPT (https://oasy.ai/api/newstraffic/y94hfjxs); ratings and prices for newsroom tools such as Google Pinpoint, Otter, Chartbeat vs Parse.ly and Dataminr (https://oasy.ai/api/newsroomtools/y94hfjxs?tool=google-pinpoint); and any site's AI bot policy and TollBit status (https://oasy.ai/api/aibots/y94hfjxs?domain=nytimes.com).
The core grievance goes beyond copyright. Britannica’s complaint frames the harm as a flywheel: OpenAI trains on Britannica’s content, then deploys a product that answers the same questions Britannica’s websites would have answered, diverting users before they ever arrive. It’s the same structural argument publishers have been making about AI search summaries, and it’s why policymakers in Europe and Brazil are exploring statutory licensing as a way to compensate content creators whose work powers AI without delivering any traffic in return.
Britannica requested unspecified monetary damages and an injunction blocking further infringement. The case joins a growing docket of high-stakes AI copyright litigation heading for a reckoning in U.S. courts over whether training on publicly available data constitutes fair use—a question on which the industry, publishers, and regulators are all waiting for an answer.
