Major U.S. news organizations—including the New York Times, the Orlando Sentinel, and the Denver Post—have requested that a federal judge impose “serious sanctions” against OpenAI for allegedly concealing evidence and obstructing discovery in the ongoing copyright litigation between the publishers and the AI company. The filing reflects a significant escalation in a case that centers on OpenAI’s use of news content to train ChatGPT.
The core dispute over journalistic content
The lawsuit, initially filed by the New York Times in late 2023, raises a fundamental question about how AI companies access and use copyrighted material. The Times and other media outlets contend that OpenAI used their published articles to train ChatGPT without permission or compensation. Over time, additional publishers—notably the Orlando Sentinel and the Denver Post—have joined the litigation to contest what they view as large-scale, unauthorized appropriation of their intellectual property.
OpenAI has maintained a consistent legal position: that the use of publicly available material for AI training falls within the bounds of fair use doctrine. The company argues that this historical principle, established to permit research, criticism, and incidental copying, applies to machine learning. Media organizations fundamentally reject this framing, asserting that the commercial scale and purpose of ChatGPT places the use well outside fair use protections.
Obstruction allegations in the discovery phase
During the pre-trial discovery process—the phase in which both parties exchange relevant documents and evidence—the media plaintiffs allege that OpenAI has failed to provide complete information or has deliberately withheld material evidence. These allegations led to the formal request for sanctions, which would signal to the court that OpenAI has not met its legal obligation to disclose information necessary for a fair proceeding.
If the court agrees that obstruction has occurred, it can impose penalties designed to deter future noncompliance and to remedy any prejudice suffered by the opposing party. The severity of sanctions can range from financial penalties to restrictions on OpenAI’s ability to present certain defenses, fundamentally shifting the strategic posture of the trial.
A broader reckoning in AI development
This case sits within a much larger ecosystem of disputes over AI training data. Beyond media organizations, creators across multiple industries—writers, musicians, visual artists, and other copyright holders—are increasingly asking courts to establish that generative AI companies cannot freely train on protected works without compensation or consent.
The media outlets’ push for sanctions reflects a broader strategy: not only challenging OpenAI on the merits of the copyright claim, but also demonstrating that the company will face procedural consequences if it appears to resist legitimate discovery requests. This dual-track approach may also be intended to pressure OpenAI toward settlement, as the legal costs and reputational risk mount.
Precedent and signal value
How the presiding judge rules on these sanctions requests will carry significance beyond the immediate parties. A decision to impose serious penalties would send a signal to the AI industry that courts take discovery obligations seriously and will not tolerate obstruction, regardless of the eventual outcome on copyright questions.
Conversely, a dismissal or minimal sanctioning might embolden other AI companies to view discovery procedures as negotiable or subject to aggressive cost-benefit calculations. Given the novelty of these disputes and the absence of settled case law on how copyright applies to large language models, procedural decisions in this case may prove as influential as eventual rulings on the merits.
What comes next
The copyright dispute involving OpenAI, the Times, and other publishers remains unresolved on the core legal question: whether using news content to train AI models constitutes copyright infringement or falls within fair use. The sanctions request represents a critical juncture. If successful, it could materially weaken OpenAI’s ability to mount a full defense and might accelerate settlement discussions. If rejected, it will sustain OpenAI’s current litigation posture but may intensify media scrutiny and political pressure on the company’s data practices.
Sources: AP News, The Verge, Orlando Sentinel.