OpenAI Counters New York Times Allegations, Claims NYT Used Deceptive Prompts to Make ChatGPT Reproduce Articles
-
OpenAI claimed in a Monday filing that The New York Times used "deceptive prompts" to make ChatGPT reproduce its content. Therefore, the company requested the U.S. Southern District of New York court to dismiss several allegations in the agency's copyright infringement lawsuit.
OpenAI stated that the Times exploited a vulnerability currently being fixed and directly input article content into the chatbot to generate outputs. The company said: "Ordinary people don't use OpenAI's products this way," citing a Times article from April 2023 titled "35 Real Ways People Are Using AI Right Now." This closely resembles the argument OpenAI made in its public response in January.
Ian Crosby, lead counsel for The New York Times, stated in an email that calling the Times' approach hacking is a misrepresentation—the Times was merely using OpenAI's products to find evidence proving they stole and reproduced the Times' copyrighted works. He added that OpenAI has not denied "copying the Times' works without permission within the statute of limitations." The New York Times sued OpenAI and Microsoft in December last year, claiming that the two companies trained their AI models using its content and that their chatbots could reproduce articles verbatim. The Times alleges that this has caused it to lose revenue and jeopardized its relationship with readers. OpenAI seeks to partially dismiss the Times' direct infringement claims regarding "copying activities conducted within three years prior to this lawsuit." Meanwhile, OpenAI has requested the court to dismiss other allegations: OpenAI's involvement in infringement; OpenAI's failure to remove infringing content; and OpenAI's creation of unfair competition through misappropriation. The Times' lawsuit also includes claims of trademark infringement, common law unfair competition misappropriation, and vicarious infringement.
At the same time, OpenAI has reduced multiple claims in the lawsuit filed by Sarah Silverman and other authors to a single direct infringement claim. Although OpenAI's request succeeded in this case, these two lawsuits are not the only ones targeting AI companies. Startups like OpenAI, Anthropic, and Stability AI are facing an increasing number of legal challenges, some from experienced and aggressive organizations with decades of experience in copyright disputes.