OpenAI has moved to dismiss portions of The New York Times’ copyright lawsuit against it, alleging that the newspaper improperly manipulated its chatbot ChatGPT and other AI systems to fabricate evidence for the case, thus breaking their terms and conditions.
In a filing submitted to a federal court in Manhattan on Monday, OpenAI argued that the Times engaged in what it termed “hacking” by inducing the technology to replicate its content through deceptive prompts that violated OpenAI’s terms of use, as reported by Reuters.
According to OpenAI, the allegations leveled by the Times fail to meet rigorous journalistic standards for accuracy. OpenAI contends that the truth, which will be revealed during the legal proceedings, is that the Times paid someone to manipulate OpenAI’s products.
In December, The Times initiated legal action against OpenAI and its primary financial supporter, Microsoft, alleging unauthorized use of millions of its articles to train chatbots. This lawsuit is part of a broader trend where copyright owners, including authors, visual artists, and music publishers, have sued tech firms over the purported misuse of their work in AI training.
Tech companies, including OpenAI, have argued that their AI systems utilize copyrighted material fairly and that such lawsuits threaten the growth of the multitrillion-dollar AI industry.
The Times’ complaint accuses OpenAI and Microsoft of attempting to exploit the newspaper’s substantial investment in journalism and create a substitute for its content. It points to instances where chatbots produced nearly identical excerpts from Times articles upon request.
In response, OpenAI stated in its filing that it required the Times tens of thousands of attempts to generate such anomalous outcomes, highlighting that ChatGPT cannot readily reproduce Times articles on demand under normal circumstances.
