DOJ Backs OpenAI in Copyright Fight, Calling AI Training 'Extraordinarily Transformative'
DOJ Backs OpenAI in Copyright Fight, Calling AI Training 'Extraordinarily Transformative'
The US Department of Justice has taken the unusual step of intervening in a closely watched copyright case, arguing that training large language models on copyrighted text should be considered fair use. The filing has intensified debate over how courts should treat AI development under existing intellectual property law, and it has prompted pushback from at least one plaintiff in the case, as well as questions from commentators about the administration's motives.
DOJ Enters OpenAI Copyright Fight With Rare Statement of Interest
On September 1, 2026, the Justice Department filed a 20-page statement of interest in the Southern District of New York, in the consolidated litigation known as In re OpenAI Copyright Infringement Litigation, which includes claims against OpenAI and Microsoft. The filing argues that training AI models on copyrighted text is “extraordinarily transformative” and satisfies the first and fourth fair use factors under copyright law.
A statement of interest is a formal but non-binding filing. It does not make the DOJ a party to the case, and it carries no formal legal authority. Even so, such filings can shape how a court frames its reasoning, and legal observers say this one could carry weight given the scope and prominence of the litigation.
The Legal Argument: Training vs. Output
Notably, the department's fair use argument is narrowly tailored to the training stage of AI development, not to what a model generates afterward. The brief does not argue that model outputs enjoy the same fair use protection, drawing a distinction between transformative internal use of copyrighted material during training and potentially infringing reproduction at the point of generation.
This distinction is significant because the underlying case is widely viewed as a bellwether for AI copyright litigation nationally. OpenAI and the New York Times, both parties to the suit, are expected to file competing summary judgment motions specifically on the fair use question, and the DOJ's intervention arrives ahead of that stage.
National Security and Competitiveness Rationale
Beyond the fair use analysis, the brief invokes national security and American competitiveness as reasons the court should weigh favorably toward AI development. The filing references a White House cybersecurity order and related directives from the Cybersecurity and Infrastructure Security Agency to support this framing, positioning the litigation's outcome as connected to broader US technological standing.
Some of the brief's factual claims about AI research progress rely on blog posts published by OpenAI and Anthropic rather than independent or peer-reviewed technical sources. Legal commentators have flagged this as a point worth scrutinizing, since the underlying factual basis for parts of the government's argument originates from the companies with a direct stake in the litigation's outcome.
Pushback From the Times and Tension With the Copyright Office
The New York Times, a plaintiff in the underlying suit, publicly criticized the DOJ's intervention the day after it was filed. The Times has framed the department's position as siding with large, well-resourced AI companies over the interests of American creators and publishers.
The department's categorical stance also sits in tension with the US Copyright Office's own prior analysis. In its 2025 report on generative AI training, the third installment of its Generative AI series, the Copyright Office reached a more cautious, fact-specific conclusion rather than a blanket determination that training qualifies as fair use. Notably, the Register of Copyrights was dismissed shortly after that report's release, a sequence of events that legal commentators have continued to describe as controversial.
Unconfirmed Conflict-of-Interest Allegations
Separately, opinion commentary has raised questions about the administration's motives in filing the brief. According to this commentary, the administration was reportedly negotiating a potential 5 percent federal equity stake in OpenAI around the same time the DOJ filed its statement of interest. This claim originates from opinion and editorial commentary rather than from primary legal filings or confirmed reporting, and it has not been corroborated by tier-one news organizations at this time.
Many observers note that even if the department's underlying legal reasoning on fair use is sound, the reported — and unconfirmed — equity discussions raise a recurring consumer and public-interest concern: whether government legal positions in high-profile tech litigation can be fully separated from the government's own potential financial interests. Readers should treat the equity stake claim as an unconfirmed allegation rather than an established fact pending corroboration from primary or regulatory sources.
What Comes Next
OpenAI and the New York Times are expected to file competing summary judgment motions addressing the fair use question directly, a step that could bring the core legal dispute to a head. Legal analysts covering the case, including commentary from firms such as Lowenstein Sandler, Skadden, and IPWatchdog, describe the litigation as a likely bellwether that could shape how courts nationally treat AI training under copyright law.
The outcome may also influence how future administrations approach similar interventions in technology-related intellectual property disputes, particularly when government statements of interest intersect with policy goals around national competitiveness and security.