According to the complaint, Britannica claims OpenAI used nearly 100,000 of its copyrighted online articles without permission to train its large language models. The publisher also alleges that ChatGPT sometimes produces responses that include full or partial verbatim reproductions of its content.

The lawsuit further argues that OpenAI improperly uses Britannica material in ChatGPT’s retrieval augmented generation system, which pulls in updated information from external sources when generating responses. In addition to copyright claims, Britannica accuses OpenAI of violating trademark law by producing inaccurate or fabricated information and falsely attributing it to the publisher, reported by TechCrunch.

Britannica says such practices harm its business by reducing traffic and revenue, as users rely on AI-generated answers instead of visiting original sources. It also warns that AI “hallucinations” could undermine public access to reliable and trustworthy information.

The case adds to a growing list of legal challenges against OpenAI. Publishers including The New York Times and Ziff Davis, along with several major newspapers across the United States and Canada, have filed similar lawsuits over copyright concerns.

A related lawsuit by Britannica against Perplexity AI is also ongoing.

There is still no clear legal precedent on whether using copyrighted material to train AI models constitutes infringement. In a separate case, Anthropic argued successfully before federal judge William Alsup that training AI on such data can be considered transformative and lawful. However, the judge ruled that Anthropic violated the law by downloading millions of books without payment, leading to a $1.5 billion class action settlement.