Source: Reuters

Reuters reports on a significant procedural ruling in the ongoing copyright litigation against OpenAI. On December 3, 2025, a federal judge denied OpenAI's motion to keep certain training logs and data records secret, ordering the company to produce them to the plaintiffs. The lawsuit, brought by a group of prominent authors, alleges that OpenAI used their copyrighted books without permission to train the Large Language Models (LLMs) that power ChatGPT.

OpenAI had argued that the training logs were irrelevant and constituted trade secrets that should be protected from discovery. However, the court found that the information was crucial for the plaintiffs to prove their claims of copyright infringement. This ruling is seen as a major victory for content creators, as it allows them to potentially uncover direct evidence of their works being ingested by the AI models, a key fact that has often been obscured by the "black box" nature of AI development.