The Intercept, Raw Story, and AlterNet filed separate lawsuits against OpenAI and Microsoft, alleging copyright infringement and removal of copyright information while training AI models. The Verge reports: According to the publication, ChatGPT “at least sometimes” reproduces “copyrighted journalistic works verbatim or nearly verbatim, without providing information such as authorship, title, copyright, terms of use, etc.” It is said that there is. According to the plaintiffs, if ChatGPT had been trained on materials containing copyright information, the chatbot “would have learned how to convey that information when responding.”
The Raw Story and AlterNet lawsuit goes further (PDF), stating that OpenAI and Microsoft argue that “If users believe that ChatGPT's responses infringe third-party copyrights, ChatGPT's popularity will decline and its revenue will be lost.'' “We had reason to know it would go down.” Both Microsoft and OpenAI provide legal protection to paying customers in case they are sued for copyright infringement due to their use of Copilot or ChatGPT Enterprise. According to the lawsuit, OpenAI and Microsoft are aware of the potential for copyright infringement. As proof of that, the publication points out how OpenAI offers an opt-out system that allows website owners to block content from their web crawlers. The New York Times also filed a lawsuit against OpenAI in December, alleging that ChatGPT faithfully reproduces journalistic works. OpenAI claims the publication exploited a chatbot bug to regurgitate the article.