The New York Times sues OpenAI, Microsoft

The complaint also reveals GPT-4's parameters: 1.76 trillion. Also, Microsoft takes 75% of OpenAI's profit until its $13 billion investment is repaid.

Deborah Yao, Editor

December 27, 2023

2 Min Read
Logo of The New York Times

At a Glance

  • The New York Times is suing OpenAI and Microsoft for copyright infringement, asking for "billions of dollars" in damages.
  • The Times said ChatGPT and Bing Chat are trained on its copyrighted content, citing examples of "near-verbatim" excerpts.
  • The lawsuit said GPT-4 has 1.76 billion parameters. Microsoft takes 75% of OpenAI's profit until its investment is repaid.

The New York Times today sued OpenAI and Microsoft for copyright infringement, citing several examples when a chatbot gave users “near-verbatim excerpts” from its articles.

According to the lawsuit, ChatGPT by OpenAI and Microsoft’s Bing Chat (renamed Copilot) were trained on “millions of articles” from the publication. These chatbots now compete with the news outlet in giving users information. Moreover, the Times alleged that Bing Chat attributed information to the Times when it was not true − a phenomenon known as hallucination.

“Defendants seek to free-ride on The Times’s massive investment in its journalism" and are “using The Times’s content without payment to create products that substitute for The Times and steal audiences away from it," according to the lawsuit.

Microsoft is a strategic partner of OpenAI, incorporating its technology into 365 productivity products and the Azure cloud platform. Microsoft has invested $13 billion in OpenAI in exchange for 75% of the startup’s profits until its investment is repaid, according to the complaint. Afterwards, Microsoft will own 49% of OpenAI.

ChatGPT can breach paywalls

The Times said it provides “trustworthy information” in a “damaged information ecosystem that is awash in unreliable content.” It is seeking “billions of dollars” in statutory and actual damages.

Related:AI News Roundup: News Content Coming to ChatGPT

The complaint reveals that GPT-4 has 1.76 trillion parameters, which OpenAI has not publicly disclosed. That compares with GPT-2’s 1.5 billion parameters, which itself was 10 times more than GPT. The lawsuit also said GPT-2 was trained on web content, including 45 million links posted by Reddit users, and this dataset prioritized quality, meaning it put Times content at the top. GPT-3 has 175 billion parameters.

The Times said users can even ask ChatGPT to breach its paywall to get at the content.

The lawsuit, filed in federal district court in Manhattan, comes after the Times said it failed to reach an agreement with OpenAI and Microsoft following talks that began in April. Axel Springer, the publisher of Politico and Business Insider, reached a licensing deal with OpenAI earlier this month while the Associated Press achieved the same in July.

Separately, the Times said that it hired an editorial director of AI initiatives to set protocols for the use of the technology in the newsroom.

Read more about:

ChatGPT / Generative AI

About the Author(s)

Deborah Yao

Editor

Deborah Yao runs the day-to-day operations of AI Business. She is a Stanford grad who has worked at Amazon, Wharton School and Associated Press.

Keep up with the ever-evolving AI landscape
Unlock exclusive AI content by subscribing to our newsletter!!

You May Also Like