API
< Back to all clusters
[TECHNOLOGY] · United States · 2 sources

started · updated

Wikimedia Foundation proposes copyright compromise in AI litigation

The Wikimedia Foundation has filed an amicus curiae brief in the United States District Court for the Southern District of New York regarding the ongoing copyright infringement litigation involving OpenAI and Microsoft. The organization seeks a middle ground in the legal battle between content publishers, such as the New York Times, and AI developers.

Wikimedia argues for a clear distinction between the training of large language models and the generation of specific outputs. The foundation suggests that pre-training models on copyrighted material should be considered fair use if the resulting model produces something new and transformative. However, it contends that when AI systems use techniques like Retrieval-Augmented Generation (RAG) to provide direct copies of material that substitute for the original work, a license should be required.

The foundation notes that while Wikimedia sites are significant sources for AI training, these systems can also divert traffic from human-maintained communities and sometimes reproduce Creative Commons content without adhering to license terms. Conversely, AI tools can assist Wikipedia editors by identifying vandalism and improving article quality.

Entities

Microsoft · New York Times · OpenAI · U.S. District Court for the Southern District of New York · Wikimedia Foundation