A prominent consortium of newspapers, anchored by the New York Times and including the New York Daily News, has escalated their copyright dispute with OpenAI by asking a federal court in Manhattan to impose sanctions against the artificial intelligence company. The newspapers contend that OpenAI made false statements to the court regarding its technical capabilities, specifically claiming it could not search its language models for evidence that it had improperly used millions of their articles to develop the systems powering ChatGPT.
The central allegation centres on what the plaintiffs describe as a deliberate misrepresentation of OpenAI's technical infrastructure. According to the newspapers' court filing, OpenAI assured the court that conducting searches of its large language models for copyrighted newspaper content would be infeasible and would burden users' privacy. Yet the newspapers now assert that OpenAI had already performed precisely such searches before any of the news publishers initiated legal action, undermining the company's claims about technological constraints.
Beyond the issue of search capabilities, the newspapers have uncovered another layer of what they characterise as obstruction. They allege that OpenAI has deliberately deleted or rendered unsearchable billions of conversations generated through ChatGPT that would contain evidence relevant to their claims. This systematic removal of data, the plaintiffs argue, represents an additional effort to obscure the scope and nature of how their copyrighted works were utilised during the development and operation of OpenAI's systems.
The legal remedy sought extends beyond a simple ruling in the newspapers' favour. The plaintiffs are requesting that the court impose financial penalties in the form of attorneys' fees and issue a finding that OpenAI's chat logs demonstrate the company misappropriated their copyrighted material. Such sanctions would carry significant weight both as a punishment and as an acknowledgment of bad faith litigation conduct.
Testing the veracity of these claims, an OpenAI employee provided testimony contradicting the company's earlier assertions to the court. According to the newspapers' documentation of this testimony, the OpenAI employee acknowledged that the company had "performed multiple searches for News Plaintiffs' content" within its systems. This admission directly contradicts OpenAI's prior representations and forms the evidentiary foundation for the sanctions motion.
The overarching litigation originates from a complaint filed by the New York Times in 2023, which characterised OpenAI's use of newspaper articles as unauthorised appropriation of intellectual property. The Times contended that OpenAI and its principal investor Microsoft deployed millions of articles to train the underlying large language model without securing permission from the publishers. This foundational claim has resonated across the media industry, triggering similar legal actions from other copyright holders.
The broader context of this dispute extends well beyond the relationship between OpenAI and traditional news organisations. Copyright owners across multiple creative sectors—including individual authors, visual artists, and music producers—have mounted concurrent legal challenges against a range of technology companies, from OpenAI to Anthropic and Meta Platforms. These cases collectively interrogate fundamental questions about intellectual property rights in the era of machine learning and whether training AI systems on copyrighted material without consent constitutes infringement.
Ian Crosby, the lead attorney representing the New York Times, issued a statement characterising OpenAI's conduct as a sustained pattern of deception spanning more than two years. Crosby alleged that OpenAI had misrepresented the facts to the newspaper consortium, to the general public, and to the judicial system itself. The attorney emphasised the contradiction between OpenAI's public stance about technical impossibility and its private operations, which apparently included routine searches of the exact type the company claimed could not be performed.
OpenAI has not yet provided a substantive response to the motion, with company representatives declining immediate comment when contacted. The company's silence on these specific allegations stands in contrast to its previous court filings, which had asserted the technical and privacy-related barriers to conducting targeted searches of its systems. How OpenAI responds to the evidence of the employee testimony and the allegations of data destruction will likely shape the trajectory of the case and influence judicial assessment of the company's credibility.
For Malaysian and Southeast Asian readers, this dispute carries implications extending beyond copyright law. The outcomes of these American cases will establish precedent for how intellectual property protections apply to artificial intelligence development globally. As countries across Asia develop their own regulatory frameworks for AI governance, the decisions made in American courts regarding copyright enforcement will inform policy discussions and litigation strategies. Furthermore, regional news organisations increasingly contribute content to international databases, potentially making them vulnerable to similar claims of unauthorised use in AI training wherever such systems are developed.
The sanctions motion represents a critical juncture in the larger copyright dispute. If the court determines that OpenAI engaged in material misrepresentation and wilful destruction of evidence, the consequences could extend far beyond financial penalties. Such findings would undermine OpenAI's credibility in ongoing litigation and potentially affect regulatory discussions about corporate transparency and accountability in the artificial intelligence sector. The case underscores mounting tensions between traditional intellectual property frameworks and the operational requirements of contemporary machine learning systems.
