The case, Bartz v. Anthropic, arose after a group of authors sued AI developer Anthropic, alleging that the company unlawfully used their copyrighted books to train its large language model, Claude. The core issue before the court was whether Anthropic’s use of these materials constituted copyright infringement or fell within the bounds of fair use under U.S. law.
Judge William Alsup ruled that Anthropic’s use of lawfully acquired and digitized books for AI training indeed qualifies as fair use. This is a historic win for AI developers, as it provides a clear judicial acknowledgment that the process of training AI models on copyrighted content—when done through proper channels and transformation—can be protected under copyright exceptions. The court recognized that AI training transforms the input data into something new, thereby aligning with the fair use principles of purpose and character of use.
However, the ruling was not an unqualified victory for Anthropic. The judge also found that the company’s retention of over 7 million pirated books in a central digital library did not qualify as fair use. This finding leaves the door open for significant legal consequences, including a potential trial scheduled for December to determine liability and damages. If found liable, Anthropic could face severe penalties, with damages of up to $150,000 per willfully infringed work.
This ruling echoes a recent decision in the Kadrey v. Meta case, where another federal judge ruled in favor of Meta’s use of certain authors’ books to train its AI model, Llama. However, unlike that case, which applied only to specific plaintiffs, the Bartz ruling provides a broader legal framework for AI training data use—though it clearly delineates limits when pirated content is involved.
The implications of this decision are wide-ranging. For AI companies, it clarifies the legal risks and provides a roadmap for how to acquire and use copyrighted materials responsibly. For authors and publishers, it underscores the importance of monitoring how their works are used in AI development and seeking remedies when unauthorized or pirated content is exploited.
Legal experts predict that this ruling will influence ongoing and future litigation in the AI space, potentially affecting major players like OpenAI and Microsoft. Moreover, the decision could guide policymakers and legislators as they consider regulations around AI and intellectual property.
