OpenAI says New York Times was in collaboration before lawsuit
Prior to the claims of copyright infringement, OpenAI disclosed discussions of cooperation with The New York Times.
In response to a lawsuit that The New York Times had started, OpenAI released a blog post on January 8th. In a complaint filed on December 27, the newspaper claimed that OpenAI had utilized millions of its articles to train chatbots, in violation of copyright rules.
But OpenAI says the NYT left out important information about what happened before the lawsuit. It is worth mentioning that OpenAI said that collaboration had been explored with the publication before the lawsuit was filed. It stated:
“It seemed like things were moving along well in our conversations with The New York Times up until our final correspondence on [Dec. 19]. The talks centered on a valuable collaboration, including ChatGPT’s real-time display with attribution, whereby both parties would benefit: the users would have access to The New York Times’ reporting, and the news organization would have a new channel to reach both their current and potential audience.”
According to OpenAI, there were concerns about regurgitation in The New York Times’ coverage, but no specific examples of this were shared. OpenAI also said that it thinks the recycled content came from articles that were reprinted on third-party websites and are now rather old.
Additionally, OpenAI claimed that the NYT willfully altered inputs to generate the recycled content. After stating clearly that such behavior is unacceptable, it detailed the steps taken to address the issue.
OpenAI has said in the past that it works with news groups. The Associated Press, Axel Springer, the American Journalism Project, and New York University are among the organizations it has partnered with, as it said.
Additionally, it claimed that training initiatives that do not comply with such agreements constitute “fair use.” Although OpenAI says it has a legal right to access and use the data in question, it has also said that it offers an opt-out mechanism based on principle. According to the report, in August 2023, the New York Times excluded itself via the opt-out method.
Also Read: North Korean Hackers Lazarus Withdraw $1.2M Bitcoin From Coin Mixer January 8, 2024