Legal Battle Over AI Training Data Invokes Copyright, Cultural and Sports Issues

On Friday, filings in a high‑profile copyright trial between The New York Times and OpenAI, with Microsoft as a co‑defendant, referenced a broad array of issues, including copyright law, the arts, and sports.

The lawsuit, originally filed in 2023 by The Times and later joined by additional news organizations, alleges that OpenAI, a prominent AI startup, and its partner Microsoft violated the publishers’ copyrights by training their AI systems on millions of newspaper articles. The Times contends that AI firms now compete with it as an information source.

The briefs submitted to the U.S. District Court for the Southern District of New York center on two main issues: whether the news articles were sufficiently transformed into a new work by AI, and whether AI‑generated content substitutes for original news pieces, thereby diminishing their value.

Friday marked the deadline for the companies to submit motions for summary judgment, which could preclude a trial. Judge Sidney H. Stein is anticipated to issue a decision in the near future.

The Times maintains that its articles were not transformed and therefore are not protected by fair‑use provisions. It further argues that OpenAI and Microsoft’s AI products function as substitutes for its reporting. Counsel for The Times alleges that the companies unlawfully copied publishers’ articles while developing their AI technologies.

The filing asserts that the future of journalism—and of responsible AI—relies on safeguarding incentives for creators of the original works essential to a thriving society.

OpenAI and Microsoft have strongly defended themselves against the allegations. In its motion, Microsoft contended that current copyright law “does not allow rights holders to block transformative technologies,” citing its own AI product, Copilot, as an example.

OpenAI, in its filing, asserted that “these technologies make the world’s facts accessible to anyone who requests them, in whatever format best suits their needs.”

The filing adds that this development represents progress, and copyright law does not impede it.

The court filings also touched on diverse subjects, including poet Lord Byron, artist Andy Warhol, and the historic New York Giants baseball team.

Microsoft’s filing observed that Byron, the 19th‑century British poet who argued that “literature and science were at odds,” was the father of Ada Lovelace, widely regarded as the world’s first computer scientist.

Lovelace viewed early computers as manifestations of “poetical science,” tools that foster progress and creativity. Microsoft argued that the AI programs central to this case exemplify that principle.

The Times referenced a precedent involving Andy Warhol, whose silk‑screened adaptation of a photographer’s image of Prince was found to infringe copyright. The Supreme Court ruled in 2023 that Warhol’s work did not qualify as fair use.

To illustrate the breadth of its technology and to reference the Southern District’s own history, OpenAI cited a web search for the New York Giants’ 1951 National League pennant—a team that ultimately lost the World Series to the New York Yankees.

The lawsuit is attracting close attention from legal scholars, news publishers, and the world’s leading technology companies.

Most AI copyright lawsuits remain in early stages and have not proceeded to trial. Only a handful have produced rulings on pivotal legal questions, according to a tracker maintained by Ed Lee, a professor at Santa Clara University School of Law.

The news organizations contend that OpenAI and Microsoft cannot demonstrate that their products do not clearly harm the news market. The Times asserts that its business suffers from markedly lower click‑through rates when search results replace links to its articles with AI‑generated summaries.

This substitution is key evidence that AI systems do not sufficiently transform the works into new expressions, the Times and other media outlets argue.

Citing expert testimony, OpenAI maintains that the plaintiffs’ businesses were not harmed because web traffic was “not negatively impacted” by ChatGPT, its widely used chatbot.

Source link