Wikimedia Warns a Bad AI Copyright Ruling Could Threaten the Open Web
In The New York Times' lawsuit against OpenAI, the Wikimedia Foundation argues that training AI models on copyrighted works should qualify as fair use—but reproducing articles verbatim should not.
Wikimedia Weighs In on a Landmark Case
As The New York Times' copyright lawsuit against OpenAI moves through the courts, the Wikimedia Foundation—the nonprofit that operates Wikipedia—has filed an amicus brief arguing that the outcome could reshape the entire open web. The Foundation contends that how courts interpret fair use in AI training will determine whether public-interest projects like Wikipedia can continue to exist alongside commercial AI systems.
The Core Argument: Training Is Fair Use, Regurgitation Is Not
The Wikimedia Foundation draws a clear line between two distinct activities:
- Training AI models on copyrighted content should be protected as fair use. This is a transformative use that creates new capabilities rather than substituting for the original work.
- Reproducing copyrighted articles verbatim should not be protected. When an AI system outputs substantial portions of a copyrighted work, that crosses into infringement.
The Foundation's position reflects a broader concern: if courts rule that merely training on publicly available content constitutes infringement, the legal and financial barriers could make it impossible for nonprofit and public-interest organizations to participate in the AI ecosystem.
Why This Matters for the Open Web
Wikipedia's content is freely licensed and publicly available precisely so that it can be reused, remixed, and built upon. The Wikimedia Foundation argues that a ruling that restricts AI training on public content could:
- Entrench incumbent AI developers by creating licensing costs that only large, well-funded companies can afford.
- Harm public-interest projects that rely on open access to information.
- Undermine the open web's foundational principle that publicly available knowledge can be used to create new works.
The 2026 Landscape
The case arrives at a critical inflection point. By 2026, courts in multiple jurisdictions are grappling with how existing copyright law applies to generative AI, and regulators in the US and EU are drafting frameworks for AI training data transparency and compensation. The outcome of the NYT v. OpenAI case—and the precedent it sets for fair use—will likely influence how these emerging rules take shape, determining not just the fate of commercial AI developers but the viability of the open knowledge ecosystem that projects like Wikipedia depend on.
The Bottom Line
The Wikimedia Foundation's brief is a reminder that the AI copyright debate extends far beyond the parties in the courtroom. How courts define fair use for AI training will shape who can build AI systems, what content they can learn from, and whether the open web remains a shared public resource or becomes a walled garden accessible only to those who can pay.
via The Verge AI