No Right to Depose Celebrity Plaintiffs in OpenAI Copyright Spat

Famous people, whether they’re celebrities or highly placed business executives, can be difficult to pin down for depositions. We’ve written about it before, in the context of the apex deposition doctrine, which shields from depositions business leaders with scant connection to the litigation.

Another ground for objecting to proposed depositions of famous would-be witnesses arises in class action litigation, where among the thousands of class members, there may very well be several well-known individuals whose testimony might be useful to opposing parties. We’re talking here about “absent class members,” persons whose interests are affected by the litigation but who are not actively involved in the case.

Federal courts are reluctant to permit depositions of absent class members—it risks turning an opt-out class action into an impermissible opt-in case.

Recently, the question of whether a handful of famous novelists could be deposed arose in In re OpenAI Inc., No. 25-md-3143 (S.D.N.Y.), a sprawling copyright infringement class action against the developer of the ChatGPT generative artificial intelligence technology. There are literally hundreds of attorneys involved in this litigation. Together, they belong to a “who’s who” of the country’s largest and most prestigious law firms.

But the plaintiffs are famous too. Cherry-picking among the plaintiffs listed on the docket sheet reveals dozens of famous authors and artists, including several big names with connections to the legal community:

• David Baldacci, lawyer and best-selling author of over 50 novels, mostly in the genre of political and law-related thrillers.
• Michael Connelly, crime beat journalist turned best-selling mystery novel author, most famous for the Lincoln Lawyer movie adaptation of his novel Blood Work.
• John Grisham, lawyer and author of The Rogue Lawyer and The Pelican Brief, has sold over 300 million copies of novels in the legal thriller genre, with many turned into successful motion pictures.
• Ayelet Waldman, lawyer, former federal public defender and novelist working in mystery genre, most well-known for seven mystery novels in the series The Mommy-Track Mysteries.
• Michael Chabon, married to co-plaintiff Ayelet Waldman, and whose mother and father were both lawyers, won the Pulitzer Prize for his 2001 novel The Amazing Adventures of Kavalier & Clay.
• Scott Turow, lawyer and best-selling novelist in legal thriller genre, most famous for the novel Presumed Innocent and his account of attending Harvard Law School in the nonfiction work One L.

These and many others, along with leading publishers and news media companies, allege that copyrights in their creative works were infringed when OpenAI used them, without permission, to develop ChatGPT.

Last month, on Oct. 8, the trial court was asked to decide whether several other famous authors could be deposed by the OpenAI. These authors – Douglas Preston, Simon Winchester, Ta-Nehisi Coates, Sarah Silverman, and Junot Díaz – were originally named plaintiffs when the lawsuit was filed. However, they subsequently withdrew as named plaintiffs and thereby acquired the status of “absent class members.”

“Absent class members,” in the argot of class action litigation, are persons who do not actively participate in the lawsuit but whose interests are represented by “named members” of the class. Only named class members are treated as parties to the litigation for procedural purposes.

Generally speaking, federal courts are reluctant to permit depositions of absent class members because it runs the risk of turning a typical “opt-out” class action into an impermissible “opt-in” class action. On the other hand, there is the risk that curtailing pretrial discovery will unfairly prevent the defendant from developing its defense of the case. So, there is a balance to be struck.

One common balancing test for deciding whether to permit discovery — including, of course, depositions — of absent class members asks whether the proposed discovery:

(1) is needed for trial preparation or to identify issues common to the class,
(2) is narrowly tailored,
(3) will impose undue burden on the absent class members, and
(4) is not available from representative plaintiffs.

OpenAI had various reasons for wanting to depose this group of authors though, generally speaking, it appeared to want to question them all about their views on generative artificial intelligence and copyright fair use. Some of them had made public statements on these matters.

In this case, the court decided, the balance should be struck in favor of those authors who did not want to sit for a deposition. OpenAI, the court said, had failed to identify any information sought that might be uniquely possessed by any of the proposed author deponents. The court reasoned that the information OpenAI was seeking could be obtained from other parties in the litigation; and it added that, if OpenAI believed it required information from these authors, it could seek to obtain it from them using less-intrusive means than a deposition.

It’s worth noting that the court’s decision regarding the absent class members poses no bar to deposing other equally famous, named plaintiff-authors in the In re OpenAI Inc. case. And these deponents, many with law degrees and litigation experience, might relish the experience. They might even use their deposition as inspiration for an upcoming novel.