A group of eight daily US newspapers sued OpenAI and Microsoft in federal court in New York on the grounds of misusing their reporters’ hard work to train the Generate AI models.
Eight more newspapers go to court
The companies illegally used millions of articles to train AI systems, including OpenAI’s ChatGPT and Microsoft’s Copilot, said Alden Global Capital’s MediaNews Group, which is the owner of the said eight newspapers.
The newspapers—The New York Daily News, The St. Paul Pioneer Press, The San Jose Mercury News, The Denver Post, The Chicago Tribune, The Orlando Sentinel, The Orange County Register, and The Sun Sentinel of Florida—are part of MediaNews Group, or subsidiaries of Alden, which is the second-largest newspaper operator in the US.
News outlets The Intercept, Raw Story, AlterNet, and the New York Times have already sued OpenAI, which is heavily backed by billions of dollars from Microsoft, and a new complaint will also follow the lawsuit.
A Microsoft spokesperson declined to comment on the subject, but an OpenAI spokesperson said that the company is careful in its design process and products, and it does support the news organization, reported Reuters.
OpenAI is seeking partnerships with news providers
The cases brought against tech companies over data usage for training generative artificial intelligence models by copyright owners like the newspaper are some of the landmark lawsuits in this domain.
Steven Lieberman, a lawyer for the MediaNews publications, told the news agency that the defendant OpenAI knows that they have to pay for expenses like worker salaries, computers, and chips, but in some way they can flee with taking content, so it must be realized that their success is built on the genuine and hard work of others.
The lawsuit claims that when OenAI and Microsoft’s AI models are prompted, they then reproduce the copyrighted content of the newspapers word for word exactly as the original. as the complaint noted,
“The current GPT-4 LLM will output near-verbatim copies of significant portions of the publishers’ works when prompted to do so.”
Source: CNBC.
It also claimed that the ChatGPT hallucinates the articles that are attributed to the newspapers, which deters their reputation, and gives examples of a fake Chicago Tribune recommendation of an infant longer that was recalled by the company due to being linked with infant deaths.
OpenAI has said that it was not previously aware of the reservations of Alden Global Capital and that they are seeking partnerships and conversations with news providers to discuss partnerships and possible solutions.