The New York Times has escalated its landmark copyright lawsuit against OpenAI and Microsoft Corp., alleging in a heavily redacted motion to its complaint that Microsoft built a supercomputing system explicitly to help OpenAI siphon copyrighted material.
The media giant’s legal broadside follows a shifting landscape in intellectual property law. In a recent ruling, the Supreme Court sided with Cox Communications Inc. against Sony Corp., establishing a higher legal bar for contributory infringement. Plaintiffs must now prove that a defendant intentionally acted to induce illegal conduct.
Recognizing a new legal precedent, the Times is seeking to align its claims against Microsoft, bolstered by what it describes as fresh evidence uncovered during the discovery process.
“Today, we asked the court for permission to file an amended complaint that further strengthens our case, clarifying our claim of contributory infringement against Microsoft based on new law and new evidence,” Graham James, a spokesperson for The New York Times, said in a statement.
James added that despite streamlining the lawsuit by voluntarily dropping claims of trademark dilution and two counts of contributory infringement, the core accusation remains unchanged: That the tech titans stole millions of articles to enrich themselves.
In its initial 2023 complaint, which marked the first major publisher lawsuit against OpenAI, the Times treated Microsoft’s supercomputing infrastructure as generic cloud services. The updated complaint now claims the “unusually complex” machine was tailor-made to train artificial intelligence (AI) models on unauthorized copyrighted works, with high-quality journalism intentionally given greater weight to allow the software to mimic sophisticated writing.
The Times further alleges Microsoft reaped astronomical financial rewards from the architecture, asserting deployment of the models “helped boost its market capitalization by a trillion dollars in the past year alone.”
To counter the tech firms’ defense that AI training constitutes “fair use,” the Times plans to leverage extensive data gathered during discovery, including user ChatGPT sessions.
The filings detail instances where users actively bypassed paywalls by asking the chatbot for the “next paragraph” of an article, as well as cases where the models spit out verbatim excerpts unprompted. The Times argues these instances serve as direct substitutes for paid subscriptions.
The amended complaint also targets “hallucinations,” citing instances where Microsoft’s Bing Chat and OpenAI’s ChatGPT fabricated quotes and attributed fake articles to the Times, including a fictitious report linking orange juice to non-Hodgkin’s lymphoma.
Microsoft and OpenAI remain defiant.
A Microsoft spokesperson dismissed the amendment as a “last-ditch effort” to rescue the case from unfavorable legal precedents. Meanwhile, OpenAI spokesperson Drew Pusateri defended the technology, saying, “Our models empower innovation, are trained on publicly available data, and are grounded in fair use.”
Legal experts note that the stakes are historically high. While OpenAI maintains its tools do not replace the newspaper, a failure to convince the court could lead to catastrophic consequences for the tech firms, including permanent injunctions, massive financial damages, or a court order requiring the companies to completely wipe their current models and start from scratch.
