Britannica and Merriam-Webster Sue OpenAI

⚡

Key Takeaways

1Encyclopedia Britannica and Merriam-Webster accuse OpenAI of using their protected content to train its AI models without permission.

2Britannica claims that OpenAI's practices, particularly through ChatGPT, harm its revenue by substituting its content.

3Other media outlets, such as the New York Times and Ziff Davis, have also filed similar lawsuits against OpenAI.

💡Why it matters — These legal actions could redefine the boundaries of using protected content in the development of artificial intelligence.

Britannica and Merriam-Webster vs. OpenAI

Encyclopedia Britannica, which owns Merriam-Webster, has filed a lawsuit against OpenAI, accusing it of “massive copyright infringement.” The complaint is based on the allegation that the AI giant used, without permission, a vast amount of protected content to train its language models. Britannica, which holds the rights to nearly 100,000 online articles, claims that this content has been exploited by OpenAI to enhance its AI systems.

Unauthorized Use of Protected Content

According to Britannica, OpenAI not only extracted these articles for training its models but also violated copyright laws by generating responses that fully or partially reproduce the original content. The publisher points to ChatGPT's retrieval-augmented generation (RAG) process, which involves scouring the web and other databases to provide updated information. Britannica also accuses OpenAI of violating the Lanham Act, a trademark law, by falsely attributing certain fabricated information to the publisher.

Impact on Revenue and Information Quality

The complaint highlights that OpenAI's practices, particularly through ChatGPT, deprive Britannica of potential revenue by providing answers that directly replace the content of publishers. Britannica also expresses concern about the consequences of these AI-generated "hallucinations," which could compromise public access to reliable and quality information.

A Growing Legal Movement

Britannica is not alone in this legal battle against OpenAI. Many other publishers and writers have also initiated similar actions. Among them, the New York Times, Ziff Davis, which owns publications like Mashable and CNET, as well as a dozen North American newspapers, such as the Chicago Tribune and the Toronto Star, have filed lawsuits for similar reasons.

Legal Precedents and Implications

While there is not yet a clear legal precedent regarding the use of protected content for training language models, a recent case has drawn attention. Anthropic, another AI company, successfully convinced a federal judge that the use of such content could be considered sufficiently transformative to be legal. However, Anthropic was ordered to pay a $1.5 billion settlement for illegally downloading millions of books, highlighting the complexity and financial stakes of these cases.

Other Ongoing Actions

In parallel, Britannica has also filed a similar lawsuit against Perplexity, another artificial intelligence company. This case is still pending, illustrating the extent of the legal challenges faced by companies using protected content to develop their technologies.