U.S. media outlets, including The New York Times (NYT), have requested that the court impose sanctions on AI company OpenAI, alleging that the firm has maliciously obstructed the legal process.
According to the NYT on July 9 (local time), 17 media organizations, including the NYT and the New York Daily News, filed a motion for sanctions with the U.S. District Court for the Southern District of New York, claiming that OpenAI failed to produce key evidence during the litigation process.
A motion for sanctions is a procedure used to request penalties, such as monetary fines or the exclusion of evidence, when an opposing party intentionally conceals evidence or violates court orders during a lawsuit.
Previously, the NYT filed a copyright infringement lawsuit against OpenAI in late 2023, alleging that the company used its articles and other content without permission to train ChatGPT.
Since then, other media outlets have filed similar lawsuits, and a federal court has been presiding over the consolidated cases since last year.
In the motion, the plaintiff media outlets argued that OpenAI refused to submit core materials that demonstrate how its AI models are trained and utilized.
They claimed that during the deposition of an OpenAI employee, it was revealed that the company actually possesses the materials it had previously claimed were impossible to produce.
The materials requested by the media outlets include training datasets and ChatGPT output logs.
The training dataset consists of the original materials used to train the AI model to understand language and generate responses, while the output logs are records of conversations between actual users and ChatGPT.
The plaintiffs argue that these materials would allow them to verify whether OpenAI actually used media content for training and to what extent ChatGPT utilized that content.
Conversely, OpenAI has countered that the media outlets are demanding personal information of users that is irrelevant to the case.
OpenAI has maintained its position that using media content for AI training falls within the scope of fair use under copyright law.
(Photo: AP, Yonhap News)