Drooid Logo
Back to story perspectives

Full Breakdown

AI Companies’ Internal Admissions Fuel Massive Copyright Lawsuit by U.S. News Publishers

By Drooid · · How we work

Background & Context

The New York Times, New York Daily News, Orange County Register and dozens of other outlets filed a federal copyright infringement suit in June 2023, alleging that OpenAI and its largest backer Microsoft used millions of articles without permission to train large-language models such as ChatGPT and Copilot. The case has expanded to represent more than 550 publications, including the Tampa Bay Times, The Austin Chronicle and the Alternative Newsweekly Foundation, highlighting a nationwide concern that AI training may erode the economic foundations of journalism.

Timeline

  • September 1 – The Trump administration submitted a brief to U.S. District Judge Sidney Stein, describing AI training as “extraordinarily” transformative.
  • September 4 – OpenAI and Microsoft presented arguments to Judge Stein emphasizing that their use of news content is a transformative fair-use practice.

Data & Statistics

  • The filings cite more than 3.9 million copies taken from The Times and 7.3 million from the Register, Daily News and sister papers for model training.
  • Internal Microsoft data show an 83-93 % drop in clicks on the newspapers’ websites after AI summaries became available.
  • Survey data indicate that 28.6 % of former subscribers are more likely to ask an AI for news, while 36 % of Times subscribers say they no longer need the newspaper after using ChatGPT.
  • OpenAI’s internal metrics recorded 1 billion monthly ChatGPT users (May 2026) and roughly 1 million weekly prompts for local news.
  • Only 1.3 % of 45,000 Copilot conversations were identified as current-affairs queries.

Official Statements & Responses

  • The same spokesperson characterized Brent Hecht’s “astonishing theft” comment as an individual employee’s perspective, not a legal analysis representing Microsoft.
  • OpenAI declined to comment on the newly released filings, and a New York Times spokesperson also declined to comment.

Conflicting Reports & Gaps

  • The Justice Department’s brief supports the tech companies, arguing that large-scale AI training provides a “transformative public benefit” that outweighs any competitive harm.
  • Newspaper plaintiffs counter that the same AI outputs “misinform users” and flood the market with low-quality “pink-slime” news, a claim not addressed in the companies’ filings.
  • The filings do not disclose the methodology used to calculate the 83-93 % click-through decline, leaving a gap in understanding the precise impact on web traffic.

Verbatim Quotes

  • “Throughout this case, defendants insisted that these documents be treated as confidential so that the public could not see them,” — Steven Lieberman, attorney

What’s Next

Judge Sidney Stein is expected to issue a summary-judgment decision that will determine whether the case proceeds to trial, potentially as early as 2027. The ruling will shape the legal standards for AI training on copyrighted material and could set precedent for future disputes involving other tech firms and content creators.