OpenAI Accuses NY Times of Misusing Chatbot in Lawsuit

OpenAI has formally requested a Manhattan federal court to partially dismiss a copyright lawsuit brought by The New York Times, alleging that the newspaper used deceptive techniques to craft evidence of copyright infringement. According to OpenAI, The Times made “tens of thousands of attempts” with “deceptive prompts” to coax ChatGPT and other AI systems into reproducing the newspaper’s copyrighted content without authorization. This action, which OpenAI labels as a violation of its terms of use, is part of a broader accusation by The New York Times and its primary financial supporter, Microsoft, of unlawfully using the newspaper’s articles to train their chatbots. This case is a snapshot of the growing legal challenges from copyright owners—including authors, visual artists, and music publishers—against tech companies over the use of copyrighted materials in AI development, spotlighting unresolved legal questions about the fair use doctrine’s applicability to AI training.

In response to these allegations, OpenAI counters that The New York Times’ investigation method, often referred to as prompt engineering or “red-teaming,” is a misuse of their products, equating it to hacking. This technique, designed to identify system vulnerabilities, was utilized by The Times to gather evidence against OpenAI, a move disputed by Ian Crosby, a representative for The Times, who defends their right to investigate the unauthorized use of their copyrighted content. Amidst this legal tussle, OpenAI defends the necessity of using copyrighted material for developing competent AI systems, arguing that limiting training data to public domain sources would hamper AI’s evolution and relevance. The company is actively seeking lawful ways to incorporate copyrighted content into their models, having already reached agreements with certain publishers and engaging in ongoing discussions with others. This legal battle underscores an urgent need for explicit guidelines and agreements that ensure AI technology’s advancement does not compromise intellectual property rights, while also fostering innovation and the widespread sharing of knowledge.