Wikimedia Urges Court to Distinguish AI Training
The Wikimedia Foundation has asked a federal court to consider how AI model training differs from systems that reproduce web content in the New York Times copyright case against OpenAI and Microsoft.
Updated on Nov. 2, 2026 in Artificial Intelligence
Live Poll
Should AI companies have to license copyrighted content before using it for training?
The Wikimedia Foundation filed an amicus brief in federal court in the Southern District of New York in The New York Times' lawsuit against OpenAI and Microsoft. The brief proposes separating AI model training from retrieval-augmented generation systems that reproduce existing web content.
Why it matters
The case could shape how courts treat the use of copyrighted and openly licensed online material to build AI systems. The foundation says it is defending the free and open human knowledge that Wikipedia depends on.
The brief cites Wikipedia's scale of more than 67 million articles in roughly 300 languages, sustained by about 250,000 monthly volunteer contributors. The court has not yet addressed whether retrieval-augmented generation falls under fair use.
The players
Wikimedia Foundation
The nonprofit organization that operates Wikipedia and advocates for free and open access to human knowledge.
The New York Times
The major American newspaper that sued OpenAI and Microsoft in 2023 over the use of its copyrighted articles.
OpenAI
The artificial intelligence research company that develops large language models and is a defendant in the lawsuit.
Microsoft
The technology company that is a defendant alongside OpenAI and is a major investor in OpenAI's AI systems.
The details
The Wikimedia Foundation's brief argues that training models on copyrighted material can qualify as fair use when the resulting models are meaningfully distinct from their source material. It contends companies should not automatically assume retrieval-augmented generation, which reproduces existing web content, is covered by the same protection.
Timeline
• 2023: The New York Times sued OpenAI and Microsoft over the use of its articles to develop AI systems.
• Friday: The Wikimedia Foundation filed its amicus brief in federal court.
Legal Landscape
The filing places Wikimedia's position within the broader fair use dispute over whether training large language models on copyrighted text is lawful under U.S. copyright law. It proposes that retrieval-augmented generation be treated separately from training, a distinction the court has not yet been asked to decide.
The outcome could determine whether AI tools that summarize or reproduce web content can be built and offered without licensing agreements. Readers of AI products may eventually see differences in how answers cite or draw from online sources.
The takeaway
The Wikimedia brief suggests that how an AI system is built, whether it learns from text or retrieves and repeats it, may matter as much as the fact that it uses copyrighted material. That distinction could influence how future AI tools are designed and licensed.
Further reading
For broader context on how courts are handling these disputes, see Artificial Intelligence coverage on Wisevoter.
Live Poll
Should AI companies have to license copyrighted content before using it for training?
