Judge orders OpenAI to release ChatGPT conversations, benefiting newspapers in ongoing copyright lawsuit.

A Manhattan judge has mandated that OpenAI supply the Daily News and various other media organizations with millions of anonymous chat logs generated by ChatGPT, amidst a significant copyright infringement case currently in progress. In an order released Wednesday, Manhattan Magistrate Judge Ona Wang denied OpenAI’s motion to reconsider her previous ruling, which required the company to relinquish 20 million output logs from its AI chatbot to the news outlets involved.

The media organizations aim to analyze a representative sample of ChatGPT’s user interactions to assess the model’s handling of journalists’ content and investigate potential copyright violations. This ruling arises from a comprehensive consolidated class-action lawsuit filed in 2023 against both Microsoft and OpenAI. The lawsuit includes allegations from prominent outlets such as The New York Times, the Daily News, as well as entities affiliated with Tribune Publishing and MediaNews Group, claiming that OpenAI has unlawfully appropriated and distorted their copyrighted materials. The Authors Guild and a variety of bestselling authors are also involved in this complex legal battle.

MediaNews Group’s Executive Editor, Frank Pine, expressed strong criticism of OpenAI’s leadership for attempting to conceal evidence regarding their alleged reliance on content from journalists. Pine voiced intentions to hold the tech co-creator accountable for its supposed ongoing misappropriation of journalistic work, emphasizing that compensation is warranted for the protected content utilized in OpenAI’s products.

In response, OpenAI has emphasized their commitment to user privacy following the court’s order. A spokesperson highlighted a blog post from company executive Dane Stuckey, which reiterated the organization’s resolve to explore all options available to safeguard user data.

Judge Wang addressed OpenAI’s privacy concerns, acknowledging that the company was nearing completion of an internal process aimed at anonymizing the chat logs in question. She noted that the safeguards in place adequately protected user privacy during this litigation. Wang concluded that the chat conversations are “clearly relevant” to claims made by the news outlets, as they may contain reproductions of copyrighted works and are essential for OpenAI’s defense strategies.

The judge underscored the proportionality of producing these logs for the case, observing that the requested 20 million logs represent a negligible fraction of the tens of billions of output logs held by OpenAI. Following the anonymization process, OpenAI will be mandated to release the logs within a seven-day timeframe. The tech company has also appealed Wang’s earlier ruling to Manhattan Federal Judge Sidney Stein, who is overseeing the overall case.

The court’s finding that OpenAI had initially withheld critical evidence has been a notable aspect of this legal proceeding. Steven Lieberman, representing MediaNews Group and Tribune Publishing, highlighted the implications of the court’s ruling regarding OpenAI’s motivations for delaying compliance with the order, suggesting it casts a negative light on the company’s intentions.