In a long-awaited statement from OpenAI following The New York Times’ explosive lawsuit against the tech company, it claimed that training AI models is fair use, which is supported by authors and creators. The organisation asserts that the lawsuit lacks merit, viewing it as an opportunity to clarify its business and technological approach.
Read: Authors launch new lawsuit against Microsoft and OpenAI
In December, the publication initiated legal proceedings against Microsoft and OpenAI, claiming infringements of intellectual property due to the inclusion of its journalistic content in the training data for ChatGPT. According to a filing in the US District Court for the Southern District of New York, the Times seeks to hold Microsoft and OpenAI accountable for “billions of dollars in statutory and actual damages” related to the “unlawful copying and use of The Times’s uniquely valuable works.”
“Defendants’ unlawful use of The Times’s work to create artificial intelligence products that compete with it threatens The Times’s ability to provide that service.”
The New York Times vs OpenAI
However, OpenAI disagreed, viewing it “as an opportunity to clarify our business, our intent, and how we build our technology.” It also accused The New York Times of “not telling the full story,” adding that the move had come as a “surprise and disappointment.”
The lawsuit filed by The Times is among several recent legal cases targeting the creators of widely-used generative AI tools like ChatGPT. In September, notable American writers, such as Jonathan Franzen, John Grisham, George R.R. Martin, and Jodi Picoult, brought a lawsuit against OpenAI for purportedly using their literary works to train ChatGPT without permission. Earlier, in July, two authors initiated a comparable legal action against OpenAI, claiming the unauthorised use of their books in the training of the company’s chatbot.
AI data training vs legality
The organisation said it stands firm on the principle that training AI models using publicly available internet materials falls under fair use, citing support from various academic and industry groups, but added that it had provided an “opt-out process.”
“The principle that training AI models is permitted as a fair use is supported by a wide range of academics, library associations, civil society groups, startups, leading US companies, creators, authors, and others that recently submitted comments to the US Copyright Office.”
OpenAI
It continued: “That being said, legal right is less important to us than being good citizens.” The response comes as multiple authors have recently launched new lawsuits against the company.
Addressing the “regurgitation” bug
The company admitted that there were “regurgitation” errors at times, or spitting out entire “memorized” parts of specific pieces of content or articles, “is a rare bug that we are working to drive to zero.” However, it put the onus on users acting responsibly.
It then appears to blame The New York Times, saying it had “intentionally manipulated prompts, often including lengthy excerpts of articles, in order to get our model to regurgitate.” As a result, it said the lawsuit is without merit.
“Still, we are hopeful for a constructive partnership with The New York Times and respect its long history, which includes reporting the first working neural network over 60 years ago and championing First Amendment freedoms.”
OpenAI
“Still, we are hopeful for a constructive partnership with The New York Times,” the company added, hoping to continue a partnership with the news organisation.
[…] Read: OpenAI refutes New York Times, upholds author fair use […]
[…] Read: OpenAI refutes New York Times, upholds author fair use […]
[…] content in training leading AI models was “impossible,” has opened the door to numerous copyright infringement lawsuits. However, recent developments suggest a shift towards a more responsible and legally compliant […]