DOJ Fair Use Position on AI Training Splits From White House Framework
The Justice Department's stance that AI training qualifies as fair use creates a policy fault line that Congress will likely have to resolve through legislation.
Bottom Line
Confirmed means supported by a primary source or strong independent sources; Disputed means credible sources conflict; Open means no solid primary source evidence either way.
- ConfirmedThe DOJ has adopted a position that AI training on copyrighted material qualifies as fair use, according to Fox News reporting citing the department's own framework submission.
- ConfirmedThe DOJ position conflicts with the White House's own stated framework on AI and copyright, creating an intra-executive branch disagreement that surfaces a legislative gap in Congress.
- OpenNo congressional committee has yet advanced a bill that directly resolves the conflict between AI training practices and existing copyright law.
- OpenThe financial terms and licensing arrangements, if any, between AI developers such as OpenAI and Microsoft and news publishers remain undisclosed in public filings.
The Department of Justice has taken the position that AI companies training their models on copyrighted material qualifies as fair use under existing law, a stance that places the department at odds with the White House's own AI policy framework, according to Fox News opinion and news reporting published November 21, 2026. This story separates documented fact from theory and opinion, with that separation stated here: the DOJ's legal position and the White House framework divergence are Confirmed from the cited reporting; the downstream congressional and commercial consequences described below are Analysis or Open where no primary source has settled them.
The DOJ position is Confirmed as a formal stance the department has communicated in legal proceedings, per the Fox News report, which names OpenAI and Microsoft as companies directly involved in active disputes with news organizations over the use of their content in AI training datasets. The specific litigation vehicles and docket numbers for those disputes were not identified in the source material and are therefore Open pending a review of federal court records.
The White House issued its own AI framework that, per the same Fox News source, does not align with the DOJ's fair use interpretation. The precise text of the White House framework has not been reproduced in the available source material, so the exact nature of the disagreement is Disputed in the sense that the full documentary record has not been placed before readers here. What is Confirmed is that the Fox News report, citing the two frameworks side by side, characterizes them as conflicting.
Analysis: When two executive branch entities hold formally different legal interpretations of the same statute, Congress faces pressure to legislate a resolution. The Copyright Act, Title 17 of the United States Code, does not explicitly address AI training, a gap that multiple legal scholars and industry groups have noted in public comment periods at the Copyright Office. The Copyright Office's own AI and copyright study, initiated in 2023, is Confirmed as an ongoing proceeding under Docket No. 2023-6, though its final recommendations had not been released as of this publication date.
The congressional angle is currently Open. No committee markup of a bill directly resolving the fair use question for AI training has been confirmed in the congressional record as of November 21, 2026. The Senate Judiciary Subcommittee on Intellectual Property and the House Judiciary Subcommittee on Courts, Intellectual Property, and the Internet have both held hearings on AI and copyright in prior sessions, per congressional records, but neither has advanced legislation to a floor vote on this specific question. As reported in 119th Congress Recesses With Key Bills Unfinished, January Deadline Looms, the 119th Congress closed its session with a range of unresolved legislative items, and AI copyright policy was among the areas without a statutory resolution.
The commercial stakes involve major technology companies and the news industry. OpenAI and Microsoft are Confirmed as parties in active copyright disputes with news groups, per the Fox News source. The New York Times filed suit against OpenAI and Microsoft in December 2023, a fact Confirmed by federal court records in the Southern District of New York, Case No. 1:23-cv-11195. The financial value of any licensing agreements between AI developers and publishers is Open because no such figures appear in public filings available to this publication.
The fair use doctrine under 17 U.S.C. Section 107 provides a four-factor test that courts apply case by case. Whether AI training satisfies that test is Disputed: some federal courts have allowed fair use defenses to proceed in AI-related cases while others have not issued definitive rulings. The DOJ's adoption of a pro-fair-use stance is a Confirmed policy position but carries no binding legal authority over courts, which will conduct their own statutory analysis.
Governor Gavin Newsom's separate proposal to bring federal AI regulation to Washington, covered previously by this publication, represents a parallel track. That proposal focused on safety and transparency standards rather than copyright, meaning the two policy streams, copyright and safety regulation, remain legislatively distinct for now. The connection between the two is Open: no bill has been introduced as of this publication date that addresses both AI safety and AI copyright in a single statutory vehicle.
What is at stake for Congress is the question of whether to codify a fair use carve-out for AI training, to require licensing and compensation for rights holders, or to leave the matter entirely to the courts. Each path carries different consequences for the technology sector, the publishing industry, and individual creators. The legislative calendar for the incoming Congress, which convenes in January 2027, has not yet been published, so committee priority assignments are Open.
The intra-executive disagreement between the DOJ and the White House on a matter touching billions of dollars in commercial activity and the structure of the domestic news industry makes congressional clarification more likely, though the timing and form of that action remain Open as of this date.
Sources Named in This Story
- Fox News: Americans don't have to choose between our AI future and ripping off copyright owners (2026-11-21)
- U.S. Copyright Office AI Study Docket No. 2023-6
- SDNY Case No. 1:23-cv-11195, New York Times v. OpenAI and Microsoft
- 17 U.S.C. Section 107, Fair Use Doctrine
- The Congressional Times: 119th Congress Recesses With Key Bills Unfinished, January Deadline Looms
Still Open
- The specific federal court docket numbers for all active litigation between AI developers and news organizations beyond the SDNY New York Times case.
What would settle it: - The exact text of the White House AI framework and the precise language that conflicts with the DOJ fair use position.
What would settle it: - Whether any congressional committee has scheduled a markup of AI copyright legislation for the 120th Congress.
What would settle it: - Financial terms of any existing licensing agreements between AI developers and publishers.
What would settle it: