The Wikimedia Foundation weighs in on the NYT's case against OpenAI (www.engadget.com)

The Wikimedia Foundation has submitted an amicus brief in support of OpenAI and Microsoft in their ongoing copyright lawsuit with The New York Times, which asserts that their use of NYT content for AI model training may violate copyright laws. The Foundation warns that extreme rulings could jeopardize the open web, emphasizing the need for a balanced approach to fair use. They argue that overly restrictive definitions could stifle innovation and lead to a closed web, diminishing the availability of secondary information sources. Specifically, the brief supports applying fair use to two AI use cases: certain model training scenarios and retrieval-augmented generation (RAG), which enhances large language model (LLM) outputs by referencing external knowledge bases. However, the Foundation draws a line against RAG that involves unattributed regurgitation of copyrighted articles. They advocate that scraping copyrighted content should be deemed fair use if it contributes to "something new and different," underscoring that Wikipedia itself is a vital resource for training LLMs. This nuanced position highlights the importance of open knowledge for the development of technologies and access to information.