Justice Department Files Brief Backing OpenAI in New York Times Copyright Case
Technology & SocietyWell-covered
|
The Facts
- The Justice Department filed a statement of interest in Manhattan federal court supporting OpenAI and Microsoft in The New York Times copyright suit.
- The filing argues that training large language models on copyrighted texts is protected by the fair use doctrine.
- The New York Times sued OpenAI and Microsoft in 2023, alleging they used millions of its articles without permission.
- Government lawyers argued that a robust domestic AI industry is a national security interest.
- The United States is not a party to the case, and the filing carries advisory rather than binding legal weight.
- The filing appears to be the first time the U.S. government has taken a position in the AI training copyright cases.
- The consolidated litigation includes roughly a dozen other plaintiffs, among them Ziff Davis, book authors and The Intercept.
- The statement of interest runs 20 pages.
Context
What is a statement of interest, and does it decide the case?
Court rules allow the Justice Department to file a statement of interest in pending litigation even when the government is not a party Washington Post. Such a filing has advisory rather than legal weight, but could bolster tech companies defending against copyright claims Aol,Guardian.
What is 'fair use' and how does it apply here?
Fair use is an exception in U.S. copyright law permitting unauthorized use of protected works in certain circumstances, such as when the use is considered transformative uol.com.br. The core dispute is whether training models like ChatGPT on news articles qualifies Yahoo! Finance,WSJ.
What did the government say the consequences of a publisher win would be?
The Justice Department wrote that "the creative possibilities and public benefits" of training on such content "far outweigh any competitive harm" Yahoo! Finance. It said a ruling for publishers could hinder AI development, limit competition among AI firms and hurt U.S. competitiveness with foreign developers CNET, and argued large publishers would disproportionately benefit from licensing requirements Nieman Lab.
Where Left and Right agree, and where they split
- Where Left and Right agree
- Both treat copyright doctrine, not national security, as what actually decides this: the twenty pages bind no one, and invoking a strategic interest settles nothing legally.
- Where Left and Right split
- Whether the story is about writers being asked to underwrite an industry without payment, or about courts applying fair use as written rather than inventing a licensing regime.
- Why they won’t converge
- The divide is about values, not facts: one side treats fair use as a settled legal rule that copyright already answers, the other treats payment to creators as the prior question no statute settles.
How left and right read it
The people who actually write the words are the ones being asked to subsidize this industry, and the government just put its thumb on that scale. Twenty pages arguing that training on copyrighted text is fair use, filed on behalf of no party to the case, against reporters, book authors and outlets like The Intercept who allege millions of articles were taken without permission. Calling that a national security interest doesn't answer the question of who gets paid. Compensate the creators first.
“In its filing, the Justice Department said that constraining the development of large language models like ChatGPT would hinder important creative and scientific progress.” — Washington Post
American AI gets built here only if courts apply copyright law as written, rather than inventing a mandatory licensing regime around it. That is why the Justice Department's twenty pages matter: not a party to the case, it argued that training on copyrighted text falls within fair use and that a strong domestic AI industry is a national security interest. The filing binds no one. Fair use already answers this.
“The Justice Department is backing OpenAI in its high-stakes legal battle with the New York Times, arguing in a court filing that using copyrighted content to train artificial intelligence models does not violate the law.” — New York Post
The receipts — all 92 sources
Wire services (15)
Independent coverage (50)
Facts first. Then every angle.
The day’s biggest stories in one short brief — the facts everyone agrees on, then the competing values behind the headlines. Free in your inbox.