Dictionary sues OpenAI | Tech Crunch

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

Britannica, which owns Merriam-Webster, holds the copyright to approximately 100,000 online articles that have been scraped without permission and used to train OpenAI’s LLMs, the publisher alleges in its lawsuit.

Britannica also accuses OpenAI of violating copyright law when it produces output that includes “full or partial verbatim reproductions” of its content, and when AI Labs uses its articles in ChatGPT’s RAG (Search Augmented Generation) workflow. OpenAI’s RAG tool is how LLM scans the web or other databases for newly updated information when responding to queries. Britannica also claims that OpenAI violated the Lanham Act, a trademark law, by generating fabricated hallucinations and falsely attributing them to publishers.

“ChatGPT starves web publishers who: [Britannica] Significantly increase your revenue by generating responses to user queries that replace or directly compete with content from publishers such as. [Britannica]Britannica also claims that ChatGPT’s illusions endanger “the public’s continued access to high-quality, trusted online information.”

Britannica is joining a number of other publishers and authors in filing a lawsuit against OpenAI over copyright issues. The New York Times, Ziff Davis (owner of Mashable, CNET, IGN, PC Mag, etc.) are suing OpenAI, as well as more than a dozen U.S. and Canadian newspapers, including the Chicago Tribune, Denver Post, Sun Sentinel, Toronto Star, and Canadian Broadcasting Corporation.

A similar Britannica lawsuit against Perplexity is still pending.

There is no strong case law establishing whether the use of copyrighted content to train LLMs constitutes copyright infringement. However, in one particular instance, Anthropic was able to convince federal judge William Alsup that its use case for using content as training data was transformative enough to be legal. But Alsup argued that Anthropic violated the law by illegally downloading millions of books instead of paying for them, and that affected authors deserve a $1.5 billion class-action settlement.

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