- Key Ruling in Meta vs. Book Authors Case: A judge sided with Meta despite better judgment as authors made wrong arguments. Meta's use of copyrighted materials to train Llama AI models was considered lawful only because authors' arguments were flawed. Authors failed to show Meta's AI threatened to dilute their markets; Meta provided expert testimony showing no discernible effect on plaintiffs' sales.
- Implications for Other Authors: The ruling only applies to the 13 authors in this case. Other authors with stronger cases of market harm may still have a chance in future Meta lawsuits. In cases like Meta's, plaintiffs often win if they have better-developed records on market effects.
- Judge's Criticism of Authors: Chhabria criticized authors for a "half-hearted" defense and noted that his opinion may not align with the reality of Llama potentially harming the book sale market.
- AI Companies' Position: AI companies may have an easier time defeating copyright claims if market dilution is a trade-off for public benefit. If authors had provided evidence of market dilution, Meta might not have won.
- Training AI vs. Teaching: Chhabria emphasized that cases will win or lose based on market harm allegations, not on the virtue of products being "transformative" uses. He criticized Judge Alsup's analogy of using books to teach children compared to using them to train AI.
- Ways for Authors to Fight AI Training: Authors have three paths to fight AI training based on market harms - claiming AI outputs regurgitate their works, pointing to the market for licensing works for training and contending unauthorized copying harms it, or arguing for indirect substitution. In the Meta case, the third argument seems more promising.
- Importance of Market Dilution: Market dilution is highly relevant in AI cases as LLMs have the potential to flood the market with competing works. Courts can't ignore this obvious way new technology can harm the incentive to create.
- Roadmap for Rights Holders: Chhabria's ruling provides a roadmap for rights holders in lawsuits against AI companies, but authors in the Meta case missed an opportunity by not emphasizing indirect substitution more strongly.
**粗体** _斜体_ [链接](http://example.com) `代码` - 列表 > 引用
。你还可以使用@
来通知其他用户。