Authors Alliance Files Amicus Brief Asking the Ninth Circuit to Review Class Certification in Bartz v. Anthropic

Yesterday, Authors Alliance filed an amicus brief, joined by EFF, ARL, ALA, and Public Knowledge, with the 9th Circuit in support of Anthropic’s petition to be allowed to appeal the class certification order issued in the lawsuit it is defending over the use of books for AI training. 

Class certification is important in this case because it means that the plaintiffs who brought the suit now have the ability to legally represent, and potentially bind, class members who are not present. Here, that covers rightsholders of some 7 million books. Class certification also puts significant financial pressure on Anthropic to settle. Seven million books multiplied by copyright’s statutory damages of up to $150,000 per work infringed – as Professor Ed Lee puts it, potential “business-ending liability.” 

We previously wrote about the many ways that class certification is problematic, including ways it does not comply with Rule 23, which exists to ensure that class actions are fair and that absent class members receive adequate due process.  

Yesterday, we asked the 9th Circuit to grant the petition to review the case because, as we explain in the brief, this case has profound implications for how authors’ works are used for training large language models, a transformative technology that will enhance creativity, advance research and learning, and expand access to knowledge. Authors Alliance has also supported authors who are engaged in text data mining and AI research, relying on the very same legal defenses that Anthropic has asserted in this case. Many Authors Alliance members are likely members of this class. We have a particular interest in ensuring that the views of those authors who are supportive of fair use, open access, and wide dissemination and use of their works are not co-opted by rightsholders who have opposing views.

Our Argument to the 9th Circuit

This is the summary of the argument we make to the court: 

This case is of exceptional importance, addressing the legality of using copyrighted works for generative artificial intelligence (“AI”), a transformative technology used by hundreds of millions of researchers, authors, and others. The district court’s rushed decision to certify the class represents a “death knell” scenario that will mean important issues affecting the rights of millions of authors with respect to AI will never be adequately resolved. With tens of billions of dollars in statutory damages in play, Anthropic will likely be “forced…to settle rather than incur the costs of defending a class action and run the risk of potentially ruinous liability,” Chamberlan v. Ford Motor Co., 402 F.3d 952, 957 (9th Cir. 2005) (cleaned up)).

Leaving this class certification in place would also mean “an unsettled and fundamental issue of law relating to class actions…is likely to evade end-of-the-case review”—specifically, whether a proposed class of rightsholders with such diversity of interests can satisfy Rule 23’s commonality requirement. Id. at 959.

Astonishingly, the district court certified this class with almost no meaningful inquiry into who the actual members are likely to be. The court did no analysis of what types of books are included in the class, who authored them, what kinds of licenses are likely to apply to those works, what the rightsholders’ interests might be, or whether they are likely to support the class representatives’ positions. The district court fell far short of the “rigorous analysis” that trial courts must undertake under Rule 23(a). Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 161 (1982).

Had the district court conducted such an analysis, it would have uncovered decades of research, multiple bills in Congress, and numerous studies from the U.S. Copyright Office attempting to address the challenges of determining rights across a vast number of books. It would have revealed how the complexity of years of publishing contracts makes ownership determinations fraught and prone to conflict. It also would have shown that rightsholders have genuine and significant differences of opinion about the legality of AI and the specific uses Anthropic has engaged in. There is no realistic pathway to resolving these issues in a common way as Rule 23(a) requires. Had the court conducted the rigorous analysis required it would have also seen that a wide diversity of authors, publishers, and rightsholders now find themselves represented by three writers whose interests are antagonistic to their own. It is precisely this scenario that Rule 23(f) is meant to avoid, and so this Court should grant Anthropic’s petition for appeal.

You can read our full brief here.