Dictionary sues OpenAI | TechCrunch

Encyclopedia Britannica and Merriam-Webster have filed a lawsuit against OpenAI, alleging in its complaint that the AI ​​giant has committed “gross copyright infringement.”

Britannica, which owns Merriam-Webster, holds the copyright to nearly 100,000 online articles, which were copied and used to train LLMs in OpenAI without permission, the publisher alleges in the lawsuit.

Britannica also accuses OpenAI of violating copyright laws when it produces output containing “full or partial verbatim copies” of its content and when the AI ​​lab uses its articles in ChatGPT’s RAG (Retrieval Augmented Generation) workflow. OpenAI’s RAG tool is how LLM scans the web or other databases for newly updated information when responding to a query. Britannica also claims that OpenAI violates the Lanham Act, a trademark law, when it generates fabricated hallucinations and falsely attributes them to the publisher.

“ChatGPT is starving web publishers like them [Britannica] generate revenue by generating responses to user queries that replace and directly compete with content provided by publishers such as [Britannica]Britannica also claims that the ChatGPT hallucination jeopardizes “the public’s continued access to high-quality, trustworthy information online.”

Britannica joins a number of other publishers and writers in pursuing legal action against OpenAI over copyright issues. The New York Times, Ziff Davis (owner of Mashable, CNET, IGN, PC Mag, and others), and more than a dozen newspapers across the United States and Canada, including the Chicago Tribune, Denver Post, Sun Sentinel, Toronto Star, and Canadian Broadcasting Corporation, have filed a lawsuit against OpenAI.

A similar lawsuit from Britannica against Perplexity is still pending.

There is no strong legal precedent determining whether the use of copyrighted content for LLM training constitutes copyright infringement. But in one specific case, Anthropic successfully convinced federal judge William Alsup that this use case — using content as training data — was transformative enough to be legal. However, Allsop said Anthropic violated the law by illegally downloading millions of books, rather than paying for them, which secured a $1.5 billion class action settlement for affected writers.

OpenAI did not respond to TechCrunch’s request for comment before publication.