POLICY · REGULATION · #2816
Wikimedia files amicus brief urging narrow fair-use rules for LLM training
The Wikimedia Foundation filed an amicus brief in the lawsuits around NYT v. OpenAI arguing that pre-training large language models on copyrighted web content should often qualify as fair use when it produces something transformative, but that models reproducing or supplanting original texts in responses should require licensing and compliance with Creative Commons terms. Wikimedia warns that an overly broad ruling for either side could both harm Wikipedia’s volunteer communities and undermine the open flow of knowledge used by AI tools.
KEY POINTS
- The Wikimedia Foundation filed an amicus brief in the lawsuits around NYT v.
- OpenAI arguing that pre-training large language models on copyrighted web content should often qualify as fair use when it produces something transformative, but that models reproducing or supplanting original texts in responses should require licensing and compliance with Creative Commons terms.
- Wikimedia warns that an overly broad ruling for either side could both harm Wikipedia’s volunteer communities and undermine the open flow of knowledge used by AI tools.
WHY IT MATTERS
The brief stakes out a middle path that could shape whether courts allow broad training on copyrighted web content while protecting original works from verbatim AI reuse, affecting AI development and the sustainability of open knowledge.