Anthropic, one of the leading generative AI firms best known for its popular language model Claude, has secured a landmark settlement. They settled that copyright suit for a near-mindboggling $1.5 billion. Combined with Georgia’s share, this settlement is the largest payout in U.S. copyright law history. In so doing, it seeks to settle the last legacy claims from a series of plaintiffs known as Authors of a subclass in the Bartz v. Anthropic AI class action suit.
Aparna Sridhar, Anthropic’s deputy general counsel, emphasized the significance of the settlement, stating, “Today’s settlement, if approved, will resolve the plaintiffs’ remaining legacy claims.” According to the lawsuit, Anthropic procured these gems by pirating millions of books stored in so-called “shadow libraries.” It alleges that the company improperly used this purloined content to train its generative AI models.
In June, Judge William Alsup issued a narrow ruling in favor of Anthropic. To recap, here’s why he claims that training artificial intelligence on copyrighted material is legal. Yet his ruling has been instrumental in creating the framework that has driven the discussion of the intersection between copyright law and AI development. Judge Alsup was clear that the trial was largely centered on the company’s supposed corporate piracy. He highlighted that the larger implications of AI training weren’t even the biggest issue.
Anthropic is just the latest of several tech giants to be sued for their use of copyrighted material to train their large language models. Other major players on this legal tussle include Meta, Google, OpenAI and Midjourney. The company has been aggressively acquiring text to train up its AI prowess. One of its flagship products, of course, is Claude.
The settlement provides a much-needed financial boost to nearly half a million authors. We want to compensate writers fairly, so each eligible writer will be awarded at least $3,000. This cash relief is lifesaving for many writers and literary workers in our community. So, artists have taken to social media to express their frustration over the misuse of their work in training AI models.
Judge Alsup’s comments further illustrate the aim behind Anthropic’s LLMs. He stated, “Like any reader aspiring to be a writer, Anthropic’s LLMs trained upon works not to race ahead and replicate or supplant them — but to turn a hard corner and create something different.” This statement has sparked discussions about the ethical implications of using existing literature to fuel AI creativity versus the potential infringement on authors’ rights.
The Bartz v. Anthropic lawsuit is a leading indicator of a growing trend of litigation against technology companies. These companies are using more and more copyrighted material to train their AI technologies. These legal battles are happening as you read this. As such, they are poised to shape the long-term foundation for how copyright law engages with artificial intelligence and creative technologies more broadly.