Anthropic has agreed to pay $1.5 billion to settle a class-action lawsuit alleging it used pirated books to train its Claude AI models. Details from the case have sparked controversy over the company's book handling practices.
TL;DR
- Anthropic settles a $1.5B copyright lawsuit, the largest in U.S. history, over allegations of using pirated books to train its AI models.
- Details from the case reveal Anthropic purchased rare books in bulk, scanned them by removing spines, and then shredded the originals.
- The settlement and book destruction practices have sparked backlash from investors and authors, highlighting ongoing tensions in AI copyright law.
What happened
Anthropic recently settled a class-action lawsuit for $1.5 billion, the largest copyright settlement in U.S. history, according to the company. The lawsuit alleged that Anthropic used hundreds of thousands of pirated books to train its Claude AI models.
Details from the case revealed that Anthropic purchased rare books in bulk, scanned them using high-speed machines that sliced off their spines, and then shredded the original copies. This practice was part of the company's process to convert physical texts into digital copies for training its AI models, as reported by Futurism.
The court found that Anthropic's actions were protected by fair use since the company was converting the original physical texts into digital copies rather than redistributing them as new copies. Services like ISBNdb offer AI companies rare books while keeping buyers' identities anonymous, per 404 Media.
Why it matters
The settlement and the controversy over book destruction practices have sparked backlash from investors and authors. Michael Burry and Elon Musk criticized Anthropic on X, with Musk directing his SpaceX AI team to preserve rare books and scan them without destroying the originals.
Venture capitalist David Sacks highlighted the hypocrisy in AI companies' stance on intellectual property, noting that while Anthropic claims the right to train on all the world's output, it objects when competitors train on its output after paying for it.
This case underscores the ongoing tensions and legal battles in the AI industry regarding copyright and the use of copyrighted materials for training AI models. The outcomes of these legal battles have been mixed, with some cases still ongoing, such as the New York Times' lawsuit against OpenAI and Microsoft, and Getty Images' legal battle with Stability AI.
Key facts
- Anthropic agreed to pay $1.5 billion to settle a class-action lawsuit alleging it used pirated books to train its Claude AI models, according to the company.
- The settlement is the largest copyright settlement in U.S. history, per the company.
- Anthropic purchased rare books in bulk, scanned them by removing spines, and then shredded the original copies, as reported by Futurism.
- The court found Anthropic's actions to be protected by fair use since the company was converting physical texts into digital copies, per 404 Media.
- Services like ISBNdb offer AI companies rare books while keeping buyers' identities anonymous, according to 404 Media.
- Michael Burry and Elon Musk criticized Anthropic on X, with Musk directing his SpaceX AI team to preserve rare books and scan them without destroying the originals.
- Venture capitalist David Sacks highlighted the hypocrisy in AI companies' stance on intellectual property, noting that while Anthropic claims the right to train on all the world's output, it objects when competitors train on its output after paying for it.
- Over the past three years, authors, news publishers, artists, and image libraries have filed dozens of lawsuits against AI companies, alleging they copied copyrighted materials without permission to train generative AI models.
Context
This controversy is part of a broader trend of legal battles in the AI industry regarding copyright and the use of copyrighted materials for training AI models. The outcomes of these legal battles have been mixed, with some cases still ongoing, such as the New York Times' lawsuit against OpenAI and Microsoft, and Getty Images' legal battle with Stability AI.
The case highlights the ethical and legal challenges faced by AI companies in acquiring and using data for training their models. It also underscores the need for clear guidelines and regulations in the AI industry to address these issues.
As AI companies continue to develop and train their models, the legal and ethical implications of their practices will remain a significant area of concern and debate.
