A group of major newspapers led by the New York Times is formally asking the federal court in Manhattan to place sanctions against OpenAI in the copyright infringement case, claiming the artificial intelligence firm knowingly misled the court on its capacity to search its own systems for evidence of the copyright infringement. According to the newspapers, OpenAI was lying about its ability to search its big language models for copyrighted content, but they also assert that it had done so, in fact, before the first lawsuit was even filed. The discovery has heightened the legal conflicts between traditional media companies and the fast-growing AI sector.
The filing by the newspapers argues that OpenAI deceived the public more than just through legal tactics, saying it actively concealed its ability to find copyrighted material in its systems. In addition to misleading investors, OpenAI eradiated billions of conversations from ChatGPT or made them inaccessible to searches, thus wiping out potential evidence of how much it had used copyrighted news articles to train its AI models, the filing said. Now the newspapers are demanding the payment of lawyers’ fees and a court determination that the chat logs OpenAI created would have revealed that it was using the copyrighted materials in a way not authorized by its owner.
Their legal battle stems from a 2023 lawsuit filed by The New York Times against OpenAI and Microsoft, the company’s biggest investor, alleging that the companies used millions of their articles without permission to train the large language model underlying ChatGPT. The case is part of a larger fight between content creators and technology companies as copyright owners in other industries, such as authors, visual artists and music labels, bring similar suits against AI developers. The verdicts of these cases could have a significant impact on the future of copyright law in the era of artificial intelligence, and could set a precedent for whether AI businesses are allowed to use copyrighted material without proper licensing and compensation, or if they are not.
The legal department of the newspapers has characterized OpenAI’s alleged actions as a “defensive action that amounted to a total deception” going beyond the courtroom. In a statement, New York Times’ top lawyer Ian Crosby said OpenAI has deceived the Times, the Daily News plaintiffs, the public and the court for more than two years. Crosby pointed out a specific issue with OpenAI: the company stated that searching for duplicated versions of the newspapers’ content from ChatGPT was impractical, cumbersome, and would infringe on users’ privacy, even as it was hiding that it had already done so. This contradiction, said the newspapers, shows that they were deliberately misleading the public and tarnishing the integrity of the legal process.
The heart of the newspaper’s case lies in a statement from an OpenAI employee who was quoted in the papers admitting that OpenAI had conducted several searches for the plaintiffs’ content, belied by the company’s previous claims that it did not have the tools to search its datasets and output logs for copyrighted material. This discovery has sparked significant concerns regarding OpenAI’s integrity and its efforts to be transparent in the discovery process. The newspapers contend that if OpenAI had acted honestly from the beginning, it should have been able to provide them with this important evidence much sooner in the case, and perhaps changed the entire course of the litigation.
The advancement marks a pivotal moment in the larger discussion on AI and copyright, moving the discussion from debates on fair use and technological progress to issues of corporate transparency and adherence to the law. On one side of the argument, the tech industry has claimed that training AI models on public material is generally protected under fair use, whereas copyright holders have pointed to two arguments: that this infringes their copyright and undermines their business models and creative incentives. However, the newspapers’ call for sanctions brings a new angle to the discussion, as it could be argued that OpenAI’s actions may have breached not only copyright regulations but also fundamental legal ethics and fairness.
The impact of this case is far reaching, and could influence future copyright and transparency practices for AI companies. A ruling in favor of the newspapers could set precedents for disclosure in AI litigation and the repercussions for companies that don’t comply with this disclosure. On the other hand, if OpenAI does manage to successfully fend off the sanctions request, it could further embolden other tech firms to take a more aggressive stance in copyright cases, making it harder for content owners to establish infringement in future litigation.
The geopolitical landscape of the conflict, meanwhile, has been influenced by wider concerns around how AI is affecting journalism and creative industries in general. The concern of journalists and content creators is that AI platforms are exploiting human content without offering adequate compensation, recognition, or credit.Journalists and content creators worry that AI companies are profiting from their work without giving them proper credit, recognition, or compensation. Adding to these worries, OpenAI has also received funding from Microsoft, a company with a massive valuation.The addition of Microsoft, one of the most valuable companies in the world, as an investor in OpenAI only serves to add to these concerns and make it seem that the competing powers between large tech companies and struggling media outlets are not even.
Meanwhile, those who believe in the development of AI say that strong copyright protections could limit the innovation needed to keep the U.S. competitive in AI. They argue that any use of copyrighted content for training AI models constitutes transformative fair use because it does not just replicate the original content but rather identifies patterns and information enabling the creation of new, useful applications. This view is gaining some momentum in Silicon Valley where many see copyright protection as a hindrance to technology development, rather than an effective protector of creative expression.
