Encyclopedia Britannica has filed a federal lawsuit against OpenAI, alleging the AI company scraped and reproduced vast amounts of its copyrighted content to train ChatGPT — and then passed off Britannica’s trusted editorial work as its own. The suit, filed in the Southern District of New York, marks another major escalation in the growing war between legacy publishers and generative AI companies over intellectual property rights.
The complaint is blunt. Britannica accuses OpenAI of copying “substantial portions” of its encyclopedia entries without permission, using that material to build and refine its large language models, and then generating outputs that closely mirror Britannica’s original text. But the lawsuit goes further than most similar cases: it also includes trademark infringement claims, alleging that ChatGPT’s responses sometimes explicitly reference “Britannica” by name, creating the false impression that the encyclopedia endorses or is affiliated with OpenAI’s products.
That trademark angle is significant. It’s one thing to argue an AI model ingested copyrighted text during training. It’s another to claim the model actively trades on your brand’s reputation when delivering answers to users. According to Slashdot, Britannica contends that ChatGPT has reproduced content nearly verbatim from its articles and attributed it to Britannica, effectively co-opting 250 years of editorial credibility.
Britannica isn’t a scrappy startup picking a fight for attention. It’s one of the oldest and most recognized knowledge brands on the planet, with a continuously updated digital encyclopedia staffed by expert editors. The company has long positioned itself as a counterweight to the open-edit model of Wikipedia, emphasizing accuracy and authority. So when ChatGPT serves up information that reads like Britannica content — or worse, cites Britannica as a source without any licensing agreement — the company sees an existential threat to its business model.
And this is precisely the tension at the heart of nearly every AI copyright case now making its way through the courts. Publishers, authors, visual artists, and musicians have all argued that generative AI companies built multi-billion-dollar products on the backs of copyrighted works, treating the open internet as a free training corpus. OpenAI has consistently countered that its use of publicly available data constitutes fair use under U.S. copyright law — a defense that remains largely untested at this scale.
The fair use question is the one that matters most. Courts will have to weigh whether training an AI model on copyrighted material is “transformative” enough to qualify, and whether the outputs compete with or substitute for the original works. Britannica’s case sharpens that second prong considerably. If ChatGPT can answer an encyclopedia-style query with Britannica-quality content, why would a user ever pay for a Britannica subscription?
OpenAI has struck licensing deals with some publishers. The Associated Press, Axel Springer, and several others have signed agreements allowing their content to be used in training. But many publishers haven’t been offered deals they consider fair — or haven’t been approached at all. Britannica appears to fall into that camp.
The lawsuit arrives at a moment when the legal walls are closing in on OpenAI from multiple directions. The New York Times sued the company in December 2023, producing striking examples of ChatGPT reproducing Times articles nearly word for word. A coalition of nonfiction authors filed a class action. Getty Images sued over image generation. Each case chips away at OpenAI’s broad fair use defense, and each forces courts to grapple with questions that copyright law wasn’t designed to answer.
Britannica’s trademark claims add a novel dimension. If a court finds that ChatGPT’s invocation of the Britannica name constitutes trademark infringement or unfair competition, it could open an entirely new front in AI litigation — one that goes beyond copyright into brand dilution and consumer confusion.
No trial date has been set. OpenAI has not yet publicly responded to the suit.
For industry professionals watching these cases accumulate, the pattern is clear. Content owners with strong brands and distinctive editorial products are increasingly unwilling to let their work be absorbed into AI models without compensation or consent. The legal outcomes here will shape how AI companies source training data for years to come — and whether the current model of scraping first and negotiating later can survive judicial scrutiny.
One thing is certain: the bill for generative AI’s training data is coming due. The only question is how large it will be.
WebProNews is an iEntry Publication