Wikipedia’s operator says a bad AI copyright ruling could threaten the open web
The Wikimedia Foundation is staking out a middle ground in a group of landmark AI copyright cases, pushing courts to legally protect large language models (LLMs) training on copyrighted work - but not chatbots regurgitating that work to users. An overly black-and-white ruling, it says, could either deprive Wikipedians of useful AI-based tools or endanger […]
The Wikimedia Foundation has weighed in on a series of lawsuits involving major publishers like The New York Times and tech giants OpenAI and Microsoft. The organization argues that a strict application of fair use doctrine in AI copyright cases could have detrimental consequences for both the open web and Wikipedia's sustainability.
The Foundation contends that training AI models on copyrighted work is fair use, but reiterating articles to users without permission is not. An overly simplistic ruling, as some publishers seek, could hinder Wikipedia's use of AI tools for improving content and could potentially "gatekeep" innovative uses of copyrighted material.
Conversely, a ruling that favors AI companies could lead to the disregard of licensing terms and fragment the dissemination of knowledge. Wikimedia suggests a middle ground, proposing that while scraping websites shouldn't be considered copyright infringement, using copyrighted works to pre-train models without a license should be deemed fair use if they result in something new and different from the original sources.
However, if models reproduce copies of material in response to user queries and replace the original work, then a license should be obtained. The organization stresses the importance of respecting licensing requirements and crediting original sources under Creative Commons licenses for any adaptations or remixes.
Written by urgent.news from The Verge's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
