The Wikipedia operator told a court that AI training on copyrighted work should be fair use, but reproducing that text should not.

The Wikimedia Foundation filed an amicus brief on Friday in the New York Times copyright lawsuits against OpenAI and Microsoft. The nonprofit, which operates Wikipedia, asked courts to protect training large language models on copyrighted work while rejecting chatbot output that reproduces that work.

Publishers including The New York Times sued OpenAI and Microsoft over AI training and output. The Wikimedia Foundation took neither side and warned that simplified fair-use rulings could let copyright holders gatekeep transformative uses or let AI companies snub licensing terms and fragment knowledge flows.

Wikimedia says its sites are a critical AI training data source while the systems drain visitors and weaken the communities that maintain those sites. The foundation argues that services such as ChatGPT reproduce Creative Commons text without following the license terms, even as language models help Wikipedia editors fight vandalism, detect copyright violations and doxxing, and find articles that need improvement.

The court must decide how carefully to separate model training from output reproduction when applying fair use. Wikimedia Foundation general counsel Stephen LaPorte says a ruling that fails that test could threaten free knowledge, open source technology, and public access to information.

What matters

• Wikimedia filed an amicus brief Friday in the New York Times lawsuits against OpenAI and Microsoft.
• Broad rulings could either block transformative uses or let AI firms ignore licensing terms, Wikimedia warns.
• Watch whether courts separate model training from output reproduction when applying fair use to generative AI.

Why it matters

Watch whether courts separate model training from output reproduction when applying fair use to generative AI.

This GenAI News article was prepared in original wording using reporting and materials published by The Verge. Source reference: https://www.theverge.com/policy/1022880/wikipedia-openai-new-york-times-copyright.

Drafted by the GenAI News review pipeline.