Wikimedia’s AI Copyright Strategy Could Reshape the Open Web

The Wikimedia Foundation is stepping into one of the most important AI copyright fights in the United States, and its position cuts straight through the usual battle lines. The nonprofit behind Wikipedia wants courts to protect AI training on copyrighted work while refusing to give chatbots permission to replace the original articles they learn from.

On October 30, 2026, Wikimedia announced that it asked the U.S. District Court for the Southern District of New York for permission to file an amicus, or friend-of-the-court, brief in In re: OpenAI, Inc. Copyright Infringement Litigation. The proposed brief supports neither party, but it argues that the court must separate the act of training an AI system from the act of reproducing copyrighted material for users.

A Landmark Case With 12 Lawsuits

The consolidated litigation brings together 12 lawsuits against OpenAI. The best-known member is The New York Times’ case against OpenAI and Microsoft, which began in 2023 after The Times alleged that the companies used Times-owned content for any purpose, including building AI training datasets, in violation of copyright law.

The court will decide whether AI companies’ use of copyrighted content to develop large language models, or LLMs, qualifies as fair use. Wikimedia says the decision will carry major influence because of the court’s jurisdiction, the parties involved, and the small number of rulings that have addressed this question so far.

The Foundation’s filing is dated October 29, 2026, and addressed to Judge Sidney H. Stein. It is a letter motion requesting leave to file, with the proposed brief attached as an exhibit. The filing follows the court’s local rules and an August 7, 2026 order governing amicus participation, as modified on September 24, 2026.

Because that order tells prospective amici to seek permission, Wikimedia did not ask the parties for consent. The proposed brief contains 5,831 words, fitting within the court’s 6,000-word limit for amicus briefs. Rhett O. Millsaps II of Lex Lumina LLP serves as counsel of record, with Stephen LaPorte as General Counsel, Jacob Rogers as Associate General Counsel, and Aly Marino as Legal Counsel for the Foundation.

Training May Be Fair Use, Substitution May Not

Wikimedia’s argument aims for a middle ground. It says using copyrighted works to pre-train LLMs without a license should qualify as fair use when the systems provide “something new and different from the original sources.” Scraping websites for that training purpose, the Foundation argues, should not automatically count as copyright infringement.

But the organization draws a firm line when an AI system pulls copies of material in response to a user’s question and delivers that material “while supplanting readers’ engagement with those original works.” In that situation, Wikimedia says the output substitutes for the original instead of transforming it enough to qualify as fair use, meaning AI companies should respect licensing requirements.

That distinction puts Wikimedia between the positions advanced by publishers and AI companies. The New York Times argues for a more expansive definition of fair use for LLMs while alleging that OpenAI infringed its copyright by scraping its website to build training datasets. OpenAI and Microsoft contend that training and deploying models based on publicly available data represents technological development covered by fair use.

OpenAI has said that regurgitating work “is a rare bug.” Microsoft said in a September filing that expert analysis found Copilot rarely reproduces full sentences from news content used to ground the model, which it said “hardly undermines the transformative purpose of LLM training.”

Wikimedia warns that both sides could push the court toward an overly broad result. The Foundation says, “Wikimedia is concerned that the parties’ positions in this case push for overbroad and simplified applications of the fair use doctrine” around AI.

• A publisher-friendly ruling could let copyright holders “gatekeep” innovative and transformative uses of their work.
• An AI-company victory could provide “a green light to snub licensing terms and fragment the flow of knowledge between human authors and audiences.”
• A blanket rule in either direction could weaken the open web and limit access to AI tools.

Why Wikipedia Sees Stakes on Both Sides

Wikimedia says its websites “have become one of the most critical sources of data for all artificial intelligence and machine learning training.” At the same time, the Foundation argues that AI systems can siphon visitors away from the communities that maintain those sites, while services such as ChatGPT reproduce text licensed under Creative Commons without following the license terms.

For Wikipedia, the relationship with LLMs is not simply hostile. These systems can help editors fight vandalism, detect copyright violations and doxxing, and find articles that need improvement. The Foundation fears that an overly restrictive fair-use rule could deprive Wikipedians of useful AI tools, while a blanket fair-use ruling could encourage publishers to lock more information behind barriers.

That shift could reduce the visibility that Wikipedia and many publishers depend on for readership and for recruiting new volunteers. In a companion post on Diff, the Wikimedia movement’s blog, the Foundation said that if AI training is not fair use, development could remain possible only for the largest corporations or the best-funded developers, leaving open-source developers unable to afford or even access the technology.

Stephen LaPorte said, “A ruling that does not apply fair use principles carefully could threaten the sustainability of free knowledge, and stifle the development of open source technologies, ultimately impacting access to information for everyone in both cases.” He added, “The court needs to understand that Wikipedia is created through a collaborative process that requires a strong foundation of open and accessible human knowledge to thrive.”

Under Wikimedia’s proposed approach, AI systems using Wikipedia information would also need to credit the original source under its Creative Commons license. That license requires “any remix or adaptations” to remain under the same license.

CEO Bernadette Meehan said the Foundation is taking a position “to defend the free and open human knowledge that Wikipedia depends on.” She described expanded access to knowledge through AI as a powerful goal, but one that must respect the people, contributors, and communities that make reliable knowledge possible.

The court’s decision will help define whether the next generation of AI grows through open access, licensing, or a line drawn between learning from knowledge and replacing it. Wikimedia is asking Judge Sidney H. Stein to recognize that those are different acts—and that the future of open knowledge may depend on treating them differently.

Stay connected via Google News

Follow us for the latest AI News