The New York Times, alongside over a dozen media organizations, has accused OpenAI of deliberately withholding critical evidence related to copyright infringement claims in federal court. The publishers allege that OpenAI failed to comply with discovery rules, acting in bad faith during litigation over the use of copyrighted materials in training AI models.
According to the court filing, OpenAI resisted producing data that would clarify what copyrighted content was included in its training datasets and how it influenced outputs from ChatGPT. The plaintiffs argue the company falsely claimed it could not search for such evidence despite having done so since before the lawsuits began. The Times and other media outlets insist the withheld information is central to proving how extensively their works were used without authorization.
OpenAI responded by dismissing the accusations as attempts to violate privacy and misrepresent the company’s compliance. A spokesperson said the allegations are unfounded and framed the dispute as part of a weakening case by the publishers. OpenAI reaffirmed its commitment to protecting user privacy and fair use principles.
This conflict dates back nearly three years to the initial suit filed by the Times against OpenAI and Microsoft, accusing them of copyright infringement tied to AI training materials. The Times’ legal action prompted additional lawsuits from journalists against several tech giants, including OpenAI and Meta, accusing them of using copyrighted content without permission to develop large language models.
These parallel cases underscore the growing legal scrutiny around the intersection of artificial intelligence and intellectual property rights. The publishers’ current request for legal sanctions seeks to compel OpenAI to disclose the relevant evidence, strengthening their claims of unauthorized copying.