Duane Morris Takeaway: This week's episode of the Class Action Weekly Wire features Duane Morris partner Jerry Maatman and special counsel Justin Donoho with their discussion of a historic settlement agreement in the intellectual property class action space that is paving the way for AI copyright infringement litigation.
Episode Transcript
Jerry Maatman: Thank you to our loyal blog readers for joining us for this week's edition of the Class Action Weekly Wire. I'm Jerry Maatman, a partner at Duane Morris, and joining me today is special counsel Justin Donoho. Thanks so much for being on the podcast.
Justin Donoho: Great to be here, Jerry. Thanks for having me.
Jerry: Today we're here to discuss a massive class action case involving Anthropic and what could be one of the largest copyright settlements in the history of American jurisprudence at a staggering $1.5 billion. Let's start with the basics. What is the case about, Justin?
Justin: Jerry, this case has everything – artificial intelligence, copyright law, alleged piracy, and a $1.5 billion settlement hanging in the balance. The plaintiffs allege that Anthropic, a leading developer of AI large language models, or LLMs, used copyrighted content from their books to train Anthropic's Claude LLMs without obtaining consent. Part of that training included Anthropic allegedly downloading over 7 million digital copies of works acquired from pirating websites. The plaintiffs claimed that Anthropic's practices violated copyright law and sought damages as well as injunctive relief. Anthropic maintained that its alleged conduct fell within the bounds of fair use and was essential for the development of competitive AI technologies. Judge William Alsup of the U.S. District Court for the Northern District of California ruled in June that while it was fair use for Anthropic to use the non-pirated copyrighted material to train Claude, at least in this case, the use of pirated works for training, if plaintiffs could prove piracy at trial, could still be copyright infringement.
Jerry: After that ruling, which I believe was one of the most few significant rulings we have seen this year to weigh in on fair use for training AI, the parties ultimately came to a settlement. And the judge issued a ruling on the plaintiff's motion for preliminary approval of a class action settlement this past week. What did Judge Alsup decide in that ruling?
Justin: Judge Alsup denied preliminary settlement approval and had what he called a list of grievances regarding the settlement. One of his key concerns was that while the parties said they had an agreement in principle, the only thing that they seemed to agree on was the $1.5 billion price tag. He questioned whether that number had asterisks attached to it, because while Anthropic allegedly retained 7 million pirated books, the settlement list only included about 465,000.
Jerry: That's quite a large gap between the parties. What would it qualify to be a book on the list?
Justin: The settlement only covers works that had ISBNs or ASINs and were registered with the Copyright Office before or shortly after Anthropic downloaded them. The judge questioned whether the list was final, and if there would be more works added. He did not want to see new claims coming out of the woodwork once Anthropic pays up. He emphasized that if Anthropic was settling for $1.5 billion, the company deserves certainty and closure. The judge also worried about other hangers-on joining the case and the impact that it would have on the posture of pending AI copyright litigation across the country.
Jerry: So where does that leave the parties and leave the lawsuit, given his ruling?
Justin: Judge Alsup denied preliminary approval but gave the parties a chance to amend the proposed settlement agreement. He ordered the parties to submit an updated list of qualified works by the 15th, a reworked claims process by September 22nd, and a new hearing is scheduled for September 25th.
Jerry: Let's talk about the broader impact of the ruling. What does this suggest for other class actions in the copyright space involving AI?
Justin: A couple of things. First, in the context of allegedly training on pirated works, the action in these types of cases is likely to be on whether the plaintiffs can meet their evidentiary burden to actually show any piracy. And then outside the context of piracy, there's still a fundamental legal question out there – can you train your AI on otherwise legally obtained copyrighted works without permission? Courts this year have been drawing different conclusions and saying, "maybe," and it really depends on how the material was sourced and used, how much, and other factors, like nature of the copyrighted work and the effect on the market. And AI just makes the stakes higher. With respect to class actions involving piracy, for example, this $1.5 billion settlement is an exponential increase from the Napster and Grokster settlements in the early 2000s for $26 million, $50 million, and those kinds of digital distribution cases since then have been relatively sparse over the years. Compared with today, we are seeing many of these AI training class actions filed. So, given this legal landscape, the Anthropic settlement is likely to push more content creators to register their works with the Copyright Office, which is a requirement to be eligible for statutory damages in these types of cases.
Jerry: So, the big takeaway, maybe, to quote Yogi Berra, is "it's not over till it's over," and there's a chance this could become the largest class action copyright AI-related settlement in the history of American jurisprudence, but the court wants more precision, more fairness, more transparency. And even if this deal goes through, it's certainly not going to be the last word in AI copyright law.
Justin: Not by a long shot. There are dozens of pending AI copyright cases. The rate of new complaints we are seeing this year seems to be increasing. Courts are coming out different ways on the fair use doctrine when there's no issue of whether piracy is involved. And eventually, we could see the Supreme Court weigh in on AI and fair use.
Jerry: Well, thank you, Justin, for your thought leadership in this area, and for breaking things down for our listeners. This is one of those legal stories that really shows how quickly the law is advancing and trying to keep up with technology and the intersection of the law, IP, and AI. And thanks to all our loyal blog listeners for being here today. We'll be watching this case closely, and bring you updates when things develop. Don't forget to subscribe to our channel, and we'll see you next time.
Justin: Thanks, Jerry, always a pleasure.
Disclaimer: This Alert has been prepared and published for informational purposes only and is not offered, nor should be construed, as legal advice. For more information, please see the firm's full disclaimer.
