Wikimedia Foundation Weighs In on NYT Lawsuit Against OpenAI, Urging a Balanced Approach

The Wikimedia Foundation, which supports Wikipedia, has filed an amicus brief (a submission from a non-party) in the copyright lawsuit brought by the US newspaper The New York Times (NYT) against artificial intelligence company OpenAI and technology giant Microsoft. The court must decide whether the companies' use of NYT-published content, including for training AI models, violates copyright law. The foundation argues that an extreme ruling in either direction could threaten the open web.

The parties' positions

The NYT says companies such as OpenAI should not have free rein over everything it publishes. OpenAI and Microsoft counter that training and deploying AI models on published data is necessary for technological development and should therefore count as fair use.

The foundation's middle path

The foundation proposes a third approach. Its brief says an overly restrictive reading of fair use could hinder tech innovation. Yet a ruling in favor of blanket fair use could produce a closed web in which knowledge is weakened by a "lack of available secondary source materials."

It recommends applying fair use to two specific cases. The first is certain model-training scenarios: scraping would be acceptable when training large language models (LLMs) that are "something new and different from the original sources." The second is retrieval-augmented generation (RAG), a technique that improves LLM output by referencing authoritative knowledge bases outside the model's training data. The foundation does not back RAG, however, when a model regurgitates existing articles without attribution.

The source notes that Wikipedia itself is regarded as one of the highest-quality datasets used to train LLMs.

Stephen LaPorte, the foundation's General Counsel, wrote that a ruling that does not apply fair use principles carefully could threaten the sustainability of free knowledge and stifle open source technology development. He added that the court needs to understand that Wikipedia is created through a collaborative process that requires a strong foundation of open and accessible human knowledge.

A summary judgment in the case is expected in the near future. It is one of more than a dozen similar cases moving through the courts, and it could set an important precedent.