Wikimedia Urges Courts to Protect LLM Training on Copyrighted Works While Safeguarding Open Web

The Wikimedia Foundation, operator of Wikipedia, filed an amicus brief in a series of high‑profile AI copyright lawsuits involving publishers such as The New York Times and AI firms OpenAI and Microsoft. The brief argues that courts should grant legal protection for large language models (LLMs) that are trained on copyrighted material, but should not extend that protection to chatbots that reproduce that material for users. Wikimedia warns that a blanket ruling in favor of publishers could allow copyright holders to "gatekeep" even transformative uses of their content, while a ruling favoring AI companies could "snub licensing terms" and fragment the flow of knowledge between authors and audiences. The foundation stresses that Wikipedia sites are a critical data source for AI training, yet the same AI systems can undermine Wikipedia’s sustainability by diverting traffic and violating Creative Commons licenses. Wikimedia’s general counsel, Stephen LaPorte, notes that careful application of fair‑use principles is essential to preserve free knowledge and open‑source technologies.