A coalition of prominent American newspapers led by the New York Times has escalated its legal battle with OpenAI by requesting that a Manhattan federal court impose sanctions on the artificial intelligence company. The newspapers, which also include the New York Daily News, filed the motion on Thursday, accusing OpenAI of providing false testimony regarding its technical capabilities to search internal systems for evidence of copyright infringement. The dispute centres on whether OpenAI improperly used millions of newspaper articles to train the underlying technology powering ChatGPT, one of the world's most widely used AI applications.
At the heart of the complaint is an allegation that OpenAI deliberately misled the court and the plaintiffs about its technical limitations. The newspapers contend that while OpenAI claimed it lacked the capability to search its large language models for copyrighted content, the company had in fact already conducted such searches well before the litigation commenced. This conduct, the plaintiffs argue, represents a fundamental breach of honesty in legal proceedings and warrants court-imposed penalties including financial compensation for legal costs and an adverse inference that OpenAI's internal communications prove it violated the newspapers' intellectual property rights.
The scale of the alleged deception extends beyond simple technical misrepresentation. According to the filing, OpenAI has also deleted or rendered unsearchable billions of ChatGPT conversations that could potentially contain evidence of the newspapers' copyrighted material being reproduced or summarised by the AI system. This alleged destruction of evidence—whether through deletion or deliberate obfuscation—compounds the original copyright infringement claims and raises questions about the company's willingness to cooperate transparently with judicial oversight. The newspapers characterise this pattern as systematic concealment designed to prevent discovery of wrongdoing.
Ian Crosby, the lead attorney representing the New York Times in this matter, issued a pointed statement articulating the strength of the plaintiffs' allegations. He emphasised that OpenAI has maintained falsehoods for more than two years, not only to the newspapers and the court but to the general public as well. The company, Crosby stated, explicitly argued that searching ChatGPT outputs for copies of the Times' and Daily News' content would be infeasible, burdensome, and would violate user privacy—contentions that are now directly contradicted by testimony from OpenAI's own employees confirming that multiple such searches had already been performed.
This contradiction became evident when an OpenAI employee subsequently testified that the company had "performed multiple searches for News Plaintiffs' content," directly undermining earlier assertions made under the auspices of legal proceedings. The inconsistency raises serious questions about whether OpenAI's initial claims were made with knowledge of their falsity or resulted from inadequate internal communication about the company's actual capabilities. Either scenario reflects poorly on OpenAI's approach to litigation and its commitment to truthfulness in legal matters.
The original lawsuit, filed by the New York Times in 2023, has become emblematic of a broader industry conflict emerging around AI training practices. The case names both OpenAI and Microsoft, which is the company's largest financial backer and investor, as defendants. The central allegation is straightforward: millions of the Times' articles were incorporated into the training datasets used to develop GPT models without authorisation or compensation. This practice raises fundamental questions about whether fair use protections that traditionally apply to journalism and research extend to the context of machine learning systems.
The conflict involving the New York Times and OpenAI is not isolated but rather reflects a pattern of litigation involving multiple categories of content creators. Authors, visual artists, and music labels have all initiated legal action against major technology companies including Anthropic and Meta Platforms, alleging similar misuse of their intellectual property for AI training purposes. These cases collectively represent a reckoning for the technology industry about the terms under which AI systems may be trained on existing creative works. The outcomes could fundamentally reshape how AI companies source training data and compensate content creators.
For Southeast Asian readers and technology stakeholders, this dispute carries significant implications. As artificial intelligence development accelerates across the region, local news organisations, authors, and creative professionals will likely face similar questions about whether their work is being incorporated into training datasets without permission. The precedents established in American courts regarding fair use, copyright protection, and corporate responsibility in the AI era will influence how these issues are addressed in Malaysia and neighbouring countries. The court's decision on whether to impose sanctions could signal broader principles about corporate accountability in emerging technology sectors.
OpenAI has not immediately responded to requests for comment regarding the motion for sanctions. The company's silence on this particular allegation contrasts with its public positioning as a responsible AI developer committed to ethical practices. The absence of a prompt response may itself influence judicial perception of OpenAI's willingness to engage seriously with the allegations. As discovery continues and additional evidence emerges through depositions and document production, the case may reveal more about how AI companies approach intellectual property compliance and internal record-keeping.
The motion for sanctions represents a critical juncture in the litigation. Beyond the underlying copyright questions, courts now must determine whether OpenAI's conduct regarding testimony and evidence warrants punitive measures. Such findings could affect not only the immediate case but also set precedent for how courts evaluate corporate behaviour in AI-related intellectual property disputes. The judge's ruling will likely influence settlement negotiations and establish baseline expectations for transparency and honesty from technology companies engaged in similar disputes.
