Wikipedia’s operator says a bad AI copyright ruling could threaten the open web

The Wikimedia Foundation has filed an amicus brief in the AI copyright lawsuits between publishers like The New York Times and OpenAI/Microsoft, arguing for a nuanced application of fair use. It warns that a ruling favoring publishers could let copyright holders block transformative uses, while a ruling favoring AI companies could allow disregard for licensing and fragment knowledge flow. Wikimedia notes that its sites are crucial training data for AI but that AI systems divert visitors and sometimes violate Creative Commons licenses by reproducing text without attribution. It proposes that scraping and training on copyrighted works should be fair use if the output is transformative, but that reproducing material to supplant original works should require a license. The foundation stresses that careful fair use principles are needed to sustain free knowledge and open source technologies.

(Source：The Verge)