Opinion: AI Copyright Dispute Needs Law, Evidence and Congressional Clarity
A congressional brief challenges the Justice Department’s national-security theory, while the court still must decide the copyright questions before it.
Three House Judiciary Democrats have asked a federal court to reject the Justice Department’s argument that national-security benefits should influence the fair-use analysis for AI training. Their filing raises a legitimate separation-of-powers concern: a broad new copyright exception should come from Congress, not implication. The brief is advocacy rather than a ruling, and the underlying infringement allegations still require case-specific evidence.
What the release says
Representatives Jamie Raskin, Jerrold Nadler and Hank Johnson filed a proposed amicus brief on October 30 in the consolidated OpenAI copyright litigation in the Southern District of New York.
The lawmakers argue that the four statutory fair-use factors do not include national security and that the Justice Department’s September statement would improperly expand copyright law without congressional authorization.
The underlying cases include allegations that OpenAI and Microsoft used journalism without permission. Those allegations have not been resolved by the lawmakers’ filing, and the court retains responsibility for applying copyright law to the evidence.
Read the official release: Ranking Members Raskin, Nadler, Johnson Urge Court to Reject DOJ’s Arguments for Giving AI Companies Unlimited, Free Use of Copyrighted New York Times Journalism to Train AI ModelsOur centrist perspective
The immediate development is procedural: members of Congress are offering the court their interpretation of a statute Congress enacted. Their institutional argument deserves consideration, but it does not determine whether any particular training use is fair.
AI development can produce public benefits, including national-security applications, while creators retain an economic interest in their work. Existing fair-use doctrine already requires attention to purpose, amount used and market harm; those questions should be addressed on the record rather than replaced by a categorical slogan.
If lawmakers believe large-scale model training requires a new licensing system, research exception or security rule, legislation would provide clearer notice and democratic accountability than an expansive position adopted during private litigation.
The tradeoffs
A narrow, evidence-based fair-use analysis can protect both innovation and creative markets, while a blanket rule in either direction could suppress useful research or weaken incentives to produce original work.
Interagency consultation may improve the government’s legal position, but consultation cannot substitute for the court’s independent judgment or for legislation when a major policy change is sought.
What to watch next
- Whether the court grants leave to file the brief and how it treats the Justice Department’s national-security argument.
- Case-specific evidence about training inputs, model outputs, licensing markets and any measurable substitution for the original journalism.
This is an editorial interpretation, not an official agency statement. Factual summaries rely on the linked release; implications and recommendations are our opinion. How we work