On Friday, Encyclopedia Britannica and dictionary publisher Merriam-Webster filed a lawsuit against OpenAI alleging that it used their copyrighted content to train its AI, then generated responses that were “substantially similar” to their content, as previously reported by Reuters.
According to Britannica, OpenAI repeatedly copied its content without permission, stating, “GPT-4 itself has ‘memorized’ much of Britannica’s copyrighted content and will output near-verbatim copies of significant portions on demand. The memorized examples are unauthorized copies that [OpenAI] used to train their models, including GPT-4.”
The lawsuit goes on to include examples of responses from OpenAI’s models side-by-side with Britannica’s text, in which entire passages appear to match word-for-word. Britannica also claims that OpenAI has been “cannibalizing” its web traffic by generating responses that “substitute, or directly compete” with Britannica’s content, rather than directing users to its website the way a traditional search engine would.
It’s the latest in a growing series of copyright lawsuits from publishers aimed at AI companies over the past several years. The New York Times has made similar claims in its ongoing lawsuit against OpenAI, including accusing the AI company of copying mass amounts of its copyrighted content. In September, Anthropic settled a class action lawsuit for using copyrighted books to train its AI models, resulting in a $1.5 billion payout to the books’ authors.