OpenAI responded on Monday to a lawsuit filed by The New York Times, stating that the legal action was baseless and that it actually supported and created opportunities for news organizations. The lawsuit revolves around the unauthorized use of published work to train artificial intelligence technologies.
The Times accused OpenAI and Microsoft of infringing on its copyrights by using millions of its articles to train A.I. technologies like the ChatGPT chatbot. OpenAI defended itself in a 1,000-word blog post, stating that it collaborates with news organizations and has partnerships with some of them, including The Associated Press. It also argued that using copyrighted works to train its technologies falls under fair use according to the law. OpenAI expressed its desire to continue collaborating with news organizations to elevate their ability to produce quality journalism through the potential of A.I.
Lindsey Held, a spokeswoman for OpenAI, declined to provide further comment. The Times was the first major American media organization to sue OpenAI and Microsoft over copyright issues related to its written works. Other groups, including novelists and computer programmers, have also filed copyright suits against A.I. companies, particularly as a result of the boom in “generative A.I.” OpenAI and other A.I. companies build this technology by feeding it enormous amounts of digital data, some of which is likely copyrighted. The blog post also addressed discussions between OpenAI and The Times about a potential partnership and stated that The Times had accused OpenAI’s technology of “regurgitating” its content, but declined to provide examples.
In a statement, Ian Crosby, an attorney for The Times, contested OpenAI’s claims and stated that using The Times’s articles to build products without permission or payment does not fall under fair use by any measure.
OpenAI admitted that its technology sometimes regurgitates articles, but referred to this as a “rare bug” that it was working to solve. The lawsuit included examples showing ChatGPT reproducing excerpts from Times articles nearly word for word. OpenAI stated that intentionally manipulating their models to regurgitate content is not appropriate and goes against their terms of use.
