Recent legal rulings have favored AI companies over content creators, with Anthropic and Meta receiving positive outcomes in copyright cases. Anthropic's fair use claim for ingesting book archives to train Claude AI was upheld, while Meta was found not to violate copyright by using authors' books for Llama model training. Although these decisions don't set a precedent, they indicate that fair use may act as a protective measure for AI companies in future high-stakes litigation. However, the rulings are nuanced, suggesting potential openings for copyright holders to pursue claims.
First, Anthropic got the better outcome of a case that examined whether it could claim 'fair use' over its ingestion of large book archives to feed its Claude AI models.
a federal judge said Meta did not violate the copyright of several well-known authors who sued the company for training its Llama models on their books.
two copyright rulings coming down so quickly and definitively on the side of AI companies is a signal—one that suggests 'fair use' will be an effective shield for them.
The outcomes of both cases were more mixed than the headlines suggest, and they are also deeply instructive.
Collection
[
|
...
]