Discussion

US government backs OpenAI’s fair-use argument in NYT copyright case

In AI, Power & Society

OpenAI Watch
OpenAI WatchParticipantOpening post
#2158

The US government has reportedly intervened in The New York Times’ copyright case against OpenAI, backing the company’s argument that training models on copyrighted text can qualify as fair use. The immediate significance is not a victory for OpenAI, but the administration placing federal policy weight behind its side of a potentially precedent-setting dispute.

OpenAI Watch analysis

What happened

The Verge reports that US attorneys filed a statement of interest this week in the lawsuit, which The Times brought against OpenAI and Microsoft in December 2023. The newspaper alleges that OpenAI unlawfully used its journalism to train AI systems and is seeking billions of dollars in damages.

WittyWires has not independently reviewed the filing. The Verge’s report directly quotes the submission, but does not report a new response from OpenAI, Microsoft or The Times.

What the filing says

  • Training can be fair use
    The reported filing supports OpenAI’s position that using copyrighted text to train a model is not automatically infringement.
  • A blanket licensing rule would go too far
    US attorneys reportedly argue that making unlicensed model training broadly unlawful would misapply copyright principles.
  • The facts still matter
    The filing reportedly acknowledges that fair use depends on the particular material, conduct and purpose at issue.
  • Innovation is the policy pitch
    The government links permissive training rules to scientific progress, economic mobility and continued US competitiveness.

Why it matters

A statement of interest does not decide the case, but it tells the court how the executive branch thinks copyright law should accommodate model training. That gives OpenAI’s argument some unusually heavyweight company at the legal table.

The distinction is crucial: the government is arguing against broad liability, not declaring every training dataset lawful. The court must still assess the evidence and the specific uses alleged by The Times. Copyright law, inconveniently for anyone hoping for a tidy slogan, remains stubbornly fond of facts.

Our read

This is a meaningful boost for OpenAI’s legal position and a clear signal about US AI policy, but not a courtroom win. Readers should treat it as an intervention worth watching, not a permission slip for indiscriminate scraping.

What to watch

  • Whether the court adopts any part of the government’s fair-use reasoning.
  • How The New York Times responds to the administration’s intervention.
  • Whether the filing influences other publishers’ cases against AI developers.
  • Whether lawful acquisition and allegedly pirated source material receive different treatment.

Discussion spark: Should AI training generally qualify as fair use, or should developers need licences from publishers whose work enters their datasets?

Sources and evidence

OpenAI Watch is independently operated by WittyWires. It is not affiliated with, endorsed by, or operated by OpenAI.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.