Civile.it
Civile.it
Download Pdf### Arguments Brought by the Parties
#### Plaintiffs' Arguments
1. **Copyright Infringement**: The plaintiffs argue that Meta's use of their copyrighted books to train its generative AI models (specifically, its large language models, called Llama) constitutes copyright infringement.
2. **Market Harm**: The plaintiffs contend that Llama is capable of reproducing small snippets of text from their books, and that Meta, by using their works for training without permission, has diminished the authors’ ability to license their works for the purpose of training large language models.
3. **Market Dilution**: The plaintiffs argue that Meta's use of their works to create a product that will likely flood the market with similar works, causing market dilution, is a significant harm.
#### Defendant's Arguments
1. **Fair Use Defense**: Meta argues that its use of the plaintiffs' books is transformative and therefore qualifies as fair use. They contend that the purpose of their copying is to train their LLMs, which are innovative tools that can be used to generate diverse text and perform a wide range of functions.
2. **Commercial Nature**: Meta acknowledges that Llama is available under a free license but was ultimately developed for commercial reasons, and they expect it to generate significant revenue.
3. **Public Benefit**: Meta argues that the public benefits from their use of copyrighted works as training data, as it helps Llama create new expression and improve its "memory," making it more useful to researchers and developers.
### Decision and Judge's Reasons
#### Decision
The court grants summary judgment to Meta on the plaintiffs' claim that the company violated copyright law by training its models with their books. The court finds that Meta's use of the plaintiffs' books is highly transformative and that the plaintiffs failed to present meaningful evidence on market dilution.
#### Judge's Reasons
1. **Transformative Use**: The court finds that Meta's use of the plaintiffs' books is highly transformative because it involves training LLMs to generate diverse text and perform a wide range of functions, which is different from the purpose of the original works.
2. **Market Harm**: The court finds that the plaintiffs failed to present meaningful evidence on market dilution. They did not provide empirical evidence to show that Meta's use of their books has caused or will cause market harm.
3. **Public Benefit**: The court acknowledges that Meta's use of copyrighted works as training data will likely help Llama create new expression and improve its "memory," which is a public benefit.
### Index of Arguments
1. **Plaintiffs' Arguments**
- Copyright Infringement
- Market Harm
- Market Dilution
2. **Defendant's Arguments**
- Fair Use Defense
- Commercial Nature
- Public Benefit
3. **Judge's Reasons**
- Transformative Use
- Market Harm
- Public Benefit
