Thanks to an unprecedented $1.5 billion class action settlement against Anthropic, nearly 500,000 authors are in line to receive payments of at least $3,000 each.
Although this settlement is the largest compensation ever awarded in a U.S. copyright case, it doesn't represent a true triumph for writers—instead, it signals yet another advantage for technology firms.
Major tech players are urgently gathering as much text as they can to train their large language models, which are behind innovative AI chat tools such as ChatGPT and Claude. These same tools, however, pose a threat to creative professions, even when their responses are bland. The more data these systems process, the more advanced they become—but after scouring almost every corner of the web, these companies are running short on fresh content to use.
This is why Anthropic, the creator of Claude, resorted to sourcing millions of books from “shadow libraries” and incorporating them into their AI's training data. The lawsuit, Bartz v. Anthropic, is one among many filed against firms including Meta, Google, OpenAI, and Midjourney regarding the legality of using copyrighted material to train AI systems.
However, writers aren't benefiting from this settlement simply because their work was used to train AI—rather, this is a costly consequence for Anthropic, which recently secured an additional $13 billion in funding, for unlawfully downloading books instead of purchasing them.
In June, U.S. District Judge William Alsup ruled in favor of Anthropic, deciding that using copyrighted works to train AI models is lawful. He reasoned that this usage is “transformative” enough to fall under fair use, a provision in copyright law that has remained unchanged since 1976.
“Just as any reader who hopes to write, Anthropic’s large language models were trained on existing works not to copy or replace them, but to diverge and produce something new,” the judge explained.
It was the act of piracy, not the AI training itself, that prompted Judge Alsup to move the lawsuit forward, but Anthropic's settlement has now made a trial unnecessary.
“If approved, today’s settlement will address the plaintiffs’ remaining claims,” stated Aparna Sridhar, Anthropic’s deputy general counsel. “We continue to focus on building reliable AI systems that empower individuals and organizations, drive scientific progress, and tackle tough challenges.”
With numerous additional lawsuits concerning AI and copyright heading to court, judges will now be able to look to Bartz v. Anthropic as a reference point. However, given the significance of these issues, future rulings may well differ.
