Wikimedia Foundation Calls for Nuanced AI Copyright Ruling in OpenAI-NYT Case
Wikimedia’s October 2026 amicus brief backs unlicensed LLM training when outputs are “new and different,” but says chatbots reproducing copyrighted text should require licenses. It also alleges ChatGPT violates Creative Commons terms.
The Wikimedia Foundation filed an amicus brief in October 2026, staking out a middle-ground position in the landmark copyright lawsuits between publishers — including The New York Times — and AI companies OpenAI and Microsoft.
The nonprofit, which operates Wikipedia, argues that courts should protect large language models (LLMs) that train on copyrighted material without a license, provided those models produce something “new and different from the original sources.” However, it draws a firm line at AI chatbots that reproduce copyrighted content in direct response to user queries, effectively substituting for the original work. Those uses, Wikimedia says, should require a license.
Wikimedia’s concern is that an overly broad ruling in either direction could cause serious harm. A sweeping win for publishers could allow copyright holders to “gatekeep” even transformative uses of their work, potentially cutting off AI tools that Wikipedia editors rely on for tasks like detecting vandalism, identifying copyright violations, and flagging articles that need improvement. A sweeping win for AI companies, it warns, could amount to “a green light to snub licensing terms and fragment the flow of knowledge between human authors and audiences.”
The foundation also alleges that services like ChatGPT already violate its rights by reproducing text licensed under Creative Commons without adhering to that license’s terms — which include crediting the original source and keeping any adaptations under the same license.
“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,” said Wikimedia Foundation General Counsel Stephen LaPorte.
The stakes are significant. Wikimedia notes that its websites have become “one of the most critical sources of data for all artificial intelligence and machine learning training,” while those same AI systems draw visitors away from Wikipedia, undermining the contributor communities that keep the site running.