There is a new development in the lawsuit filed by the New York Times for infringement by the New York Times: OpenAI has asked a federal judge to dismiss some of the plaintiffs' claims, saying that the plaintiffs hired computer "hackers" to manipulate OpenAI's generative AI chatbot ChatGPT and other artificial intelligence systems to generate "misleading evidence".
According to a number of foreign media 27**, OpenAI submitted documents to the Federal Court for the Southern District of Manhattan, New York, saying that the New York Times paid hackers to manipulate OpenAI products and used "deceptive prompts that blatantly violate OpenAI's terms of use" to induce ChatGPT to generate content that copies the copyrighted works of the New York Times and keep them as evidence of infringement.
"Under normal circumstances, people would not be able to use ChatGPT to access New York Times articles at will," the document explains. OpenAI found that the New York Times had "tens of thousands of attempts to produce highly perverse results," and that the accusations were therefore contrary to its "notoriously strict journalistic standards."
On December 27 last year, the New York Times accused OpenAI and its investor Microsoft Corp. of using millions of New York Times articles to train chatbots such as ChatGPT without authorization to create alternatives to the New York Times, competing with "high-quality" products against the originals.
Earlier this year, OpenAI countered that the publicly available network material used to train AI systems is protected by relevant provisions of copyright law, and that verbatim copying is a "rare glitch" caused by improper use.
In a filing on the 26th of this month, OpenAI also said: "The New York Times cannot prevent AI models from gaining knowledge about facts. ”
However, OpenAI did not mention the name of the hacker "gunman" hired by The New York Times, nor did it identify the company as violating any anti-hacking laws.
New York Times lawyer Ian Crosby issued a response statement on the 27th, saying that the ** company "just used OpenAI's products to find evidence that it stole and copied the copyright work of the New York Times", but OpenAI "bizarrely misunderstood this as a hacking act".
According to Reuters, in addition to The New York Times, several other copyright owners, including writers, visual artists and publisher groups, have also filed lawsuits against some technology companies that develop artificial intelligence, accusing them of misusing copyrighted works in AI training. Defendants claim that their AI systems make fair use of copyrighted material, and that such lawsuits are hindering the development of this emerging industry of AI.
The relevant U.S. courts have yet to rule on the key issue of whether AI training complies with "fair use" under copyright law. Judges have so far dismissed a number of similar infringement lawsuits, citing a lack of evidence to prove that AI-created content is similar to copyrighted works.
**: CCTV news client
Process edit: u060