What happened
The New York Times, Daily News, and other media outlets requested a federal judge impose sanctions on OpenAI, alleging discovery misconduct in an ongoing copyright infringement lawsuit. Plaintiffs claim OpenAI is hiding evidence, including datasets and ChatGPT logs, that would show its AI models were trained on copyrighted news content. OpenAI denies the allegations, stating it protects user privacy and adheres to fair use principles. This action escalates a lawsuit filed in late 2023, which challenges how OpenAI and Microsoft built AI technologies using millions of news articles.
Why it matters
Legal teams face heightened scrutiny over AI training data provenance, as sanctions could compel OpenAI to disclose specific datasets and logs, setting a precedent for transparency in model development. The New York Times has spent over $28 million on AI litigation, underscoring the financial stakes. This follows Anthropic's $1.5 billion settlement for training its Claude chatbot on pirated works, indicating significant financial risk for AI developers and potential constraints on future training practices.
