Separating AI hype from reality.

DOJ Breaks With White House, Sides With OpenAI on AI Training Fair Use in Copyright Fight

November 20, 2026 · Lawsuits · OpenAI · Microsoft

The Department of Justice filed a statement of interest on September 1 supporting OpenAI and Microsoft in the New York Times' copyright lawsuit, embracing a broad fair-use defense for AI training that contradicts the White House's own March 2026 AI framework and caught the Copyright Office and Patent and Trademark Office off guard.

The Times and other publishers allege OpenAI and Microsoft used millions of articles without permission to train large language models. The tech companies argue the practice is fair use. In its brief, the DOJ told the court that "the United States has a strong interest in . . . rejecting any argument that training [large language models] on copyrighted texts violates copyright law," according to a Congressional Research Service legal sidebar that quotes the filing [2].

That position goes further than the White House's National Policy Framework for Artificial Intelligence, published in March. The framework stated that "although the [Trump] Administration believes that training of AI models on copyrighted material does not violate copyright laws, it acknowledges arguments to the contrary exist and therefore supports allowing the Courts to resolve this issue" [2].

The Carolina Principles for Emerging Technologies, adopted by G20 ministers at the Commerce Department's Innovation Ministerial, also struck a more cautious tone, recognizing "complex legal and policy questions" and the "essential role" of copyright protections in supporting creators [1].

Congressional Democrats are now demanding answers. A letter circulated this week by Rep. Chellie Pingree (D‑Maine) and signed by six other House Democrats says they are "alarmed" the Copyright Office and USPTO—agencies legally required to provide impartial input on copyright matters—were not consulted before the DOJ filing [3]. The lawmakers argue the intervention "stands to undermine public trust in the impartiality of the DOJ and its ability to represent the best interests of the nation" and ask for all draft documents and a list of any outside companies that influenced the position.

Axios reported that no career antitrust attorneys signed the September filing alongside senior DOJ officials, a departure from typical practice for statements of interest [3]. The Justice Department has not responded to requests for comment.

The split reflects a deeper policy tension. The DOJ brief argues licensing costs would burden small AI developers and threaten U.S. AI leadership and national security [1]. Critics, including former USPTO deputy director Coke Morgan Stewart writing in Fox News, counter that large AI firms are best positioned to pay and that licensing markets are already emerging [1]. Several AI startups have said paying for training data reduces legal uncertainty and improves models [1].

Courts have begun ruling on fair use in AI training, with mixed results. In June 2025 a California federal judge held that Anthropic's training on books was fair use, while a different judge in the same district ruled for Meta on similar facts but warned factor four—market harm—might weigh "decisively" against fair use in other cases [2]. In September 2026 the Third Circuit ruled against fair use for a non‑generative AI system that copied Westlaw case summaries [2].

The Copyright Office's 2025 pre‑publication report concluded "some uses of copyrighted works for generative AI training will qualify as fair use, and some will not" [2].

Legislative proposals include the bipartisan CLEAR Act (S. 3813), introduced by Sens. Adam Schiff (D‑Calif.) and John Curtis (R‑Utah), which would require AI developers to disclose copyrighted works in training datasets without resolving the fair‑use question [1][2].

With Democrats poised to take control of both chambers in January, the Pingree letter signals the first of what could be extensive oversight. House Oversight Committee incoming chair Robert Garcia has already sent document‑preservation letters to Trump administration officials, including Attorney General Todd Blanche, laying groundwork for investigations [3].

The DOJ's statement of interest is not binding but carries weight with the court. How the judiciary reconciles the competing fair‑use analyses—and whether Congress steps in—will shape the economics of generative AI for years to come.

Sources

  1. DOJ sides with OpenAI in AI copyright fight against news organizations (foxnews.com)
  2. Generative Artificial Intelligence and Copyright Law (congress.gov)
  3. Scoop: Democrats probe DOJ's role in OpenAI copyright fight (yahoo.com)
  4. BERKELEY TECHNOLOGY LAW JOURNAL (btlj.org)
  5. Americans don’t have to choose between our AI future and ripping off copyright owners - AOL (aol.co.uk)
  6. Disney Enterprises Inc. v. Midjourney Inc., 2:25-cv-05275 – CourtListener.com (courtlistener.com)