Bartz v. Anthropic
回到2025年6月23日，加州北區聯邦法院作出即決判決，判決Anthropic的生成大型語言模型對人工智能的訓練是非常具轉換的，因此符合合理使用。隨然法院作出這判決，但該判決仍留下諸多伏筆及未決問題。2025年9月雙方宣布以15億美元達成和解，Anthropic就所剽竊自Library Genesis與 Pirate Library Mirror等482,460本書，每本將近3000美元計。
It was a landmark moment that sent shockwaves through the AI copyright litigation world, where many AI developer defendants have been alleged or admitted to using training material from the same piracy-laden datasets. Copyright Alliance CEO Keith Kupferschmid stated at the time that “[w]hile the settlement amount is very significant and represents a clear victory for the publishers and authors in the class, it also proves what we have been saying all along—that AI companies can afford to compensate copyright owners for their works without it undermining their ability to continue to innovate and compete.”
Kadrey v. Meta
Just two days after the Bartz summary judgment order, on June 25, another district court in the Northern District of California issued an order on summary judgment in a case brought by book authors against Meta, finding that the use of the plaintiffs’ books for training Meta’s LLM was “highly transformative” and qualified as fair use. While the holding appeared to be detrimental to the plaintiffs in the case, the overall impact was questionable because the court made abundantly clear that its decision was very narrow due to it being based on a lack of evidence presented by counsel.
What will likely prove to be the most important aspect of the decision was the court’s thoughtful and lengthy discussion of the indirect substitutional impacts that could harm the copyright owners’ actual and potential markets. Unlike Bartz, the Kadrey court understood and appreciated that the advent of generative AI technology and the scale of its impact on copyright’s incentives to create and distribute new copyrighted works for the public to enjoy requires a more thoughtful approach that is more consistent with the spirit and purpose of copyright and the fair use defense.
While the summary judgment order was a setback for the plaintiffs, the case is proceeding on the question of whether Meta simultaneously uploaded copyrighted works while using BitTorrent technology to download them—a process known as seeding. If the court finds that Meta distributed massive amounts of pirated copyrighted works, it could be held liable for staggering damages (similar to Anthropic) and perhaps another big settlement will be reached in 2026. But what may be the most important aspect of the case is the roadmap to future copyright-owner plaintiffs on how to prevail in their AI infringement cases that was explicitly provided by the judge in the case.
