By Starr Vijolan, a senior at Ramapo College of New Jersey who served as an intern at Washington Legal Foundation during the fall of 2023 through the Washington Internship Institute.

* * * *

The class action, which was initially designed to promote judicial economy and facilitate the redress of relatively small harms, has become an industry unto itself.  Like in most industries today, artificial intelligence is beginning to play an increasingly prominent role in the class-action-litigation process. One litigation vendor asserts that its AI-driven application, ClaimClam, can revolutionize the identification of class members eligible to make claims on a class-action settlement. Recent judicial scrutiny of ClaimClam, however, raises important questions about the application’s utility.

Litigation procedural rules allow one or several plaintiffs to represent the interests of dozens or even millions of other “absent” class members that allegedly suffered the same harm. Since those individuals have no direct role in the lawsuit (and may be completely oblivious to it), the presiding court must protect their interests. Such protection is especially vital if, as most often happens with such litigation, the parties to a class action settle. In the instance of a settlement, the court must protect the due-process rights of absent class members, including those plaintiffs’ ability to opt out and file their own claims.

Thus, the process of notifying absent class members, which a court-appointed administrator manages, is critical. That process not only must effectively identify individuals with legitimate claims on the settlement but must also preserve settlement funds for injured parties by rejecting bogus claimants.

ClaimClam contends that it can enhance the notice process. Based in New York, ClaimClam uses AI to seek out class members and then files for settlement proceeds on their behalf. ClaimClam of course charges for its service—15% of the settlement award. That creates an incentive for ClaimClam to produce as many claims as possible, whether they’re legitimate or not. The service embraces and advances the plaintiffs’ bar’s view of class actions on its website, declaring that it’s an easy tool to help harmed individuals retrieve what they are “owed from naughty corporations.” The service is “always monitoring new class actions” and stands ready to “handle the boring forms.” All one needs to do is take 15 minutes to fill out a form. According to the founder, ClaimClam consumer class action claims rate, which ranges from 4% to 9%.

ClaimClam’s participation in two prominent class-action settlements has ultimately been unsuccessful. ClaimClam compiled and submitted claims against the $725 million settlement from litigation over Facebook’s Cambridge Analytica controversy. The federal court’s claims administrator rejected the bundle of ClaimClam requests because the claims lacked vital identifying information. The trial judge upheld the administrator’s decision.

ClaimClam also appeared in hearings on behalf of class members in a $255 million settlement of consumer class actions against JUUL Labs. The service submitted claims for approximately 30,000 individuals. The claims administrator rejected those submissions, reasoning, “Specific to this case, Class Counsel point to evidence that the information provided by ClaimClam to Class Members about the JLI Class Settlement and claims process was incorrect or potentially misleading.” At the final settlement hearing, U.S. District Court for the Northern District of California Judge William Orrick rejected ClaimClam’s objection. Judge Orrick stated:

The Court takes these steps to ensure that class members’ due process rights are fully protected. Allowing en masse submissions by claims aggregators like ClaimClam raises real risks that Class Members will not receive accurate information about the scope of the claim and the claims process. Allowing a third-party to submit hundreds of thousands of aggregated claims also hinders the ability of the Court-appointed Claims Administrator to communicate directly with claimants and conduct required follow up to identify fraudulent claims or verify the accuracy of claims and to resolve claim disputes (e.g., confirm hours worked in wage and hour suits, or the amount purchased in consumer suits).

Although the service cheerleads for the class-action device, some lawyers who specialize in such lawsuits have not embraced ClaimClam. One plaintiffs’ lawyer in the JUUL lawsuit labeled the service’s involvement in the settlement the “ClaimClam calamity.” Lawyers involved in class-action settlements don’t want illegitimate, AI-recruited claimants eroding the amount of money available for legitimate claimants.

Plaintiffs’ attorneys should also object to claimant-recruitment services like ClaimClam because they help validate businesses’ criticisms of the class-action device. Each failed ClaimClam participation provides added support for class-action skeptics’ argument that such lawsuits are rife with uninjured plaintiffs. And the 15% cut ClaimClam requires of each client furthers the impression that class actions elevate the quantity of claims over the quality of claims.

Technology such as artificial intelligence will take on an increasing role in a litigation process that becomes more complex each year. Judges and other class-action litigation participants have justifiably questioned whether online claims-aggregation vendors are a welcome part of that evolution. Early indications are that ClaimClam operates in a manner that counteracts its goals of advancing class actions and holding “naughty corporation” accountable. Claims administrators and courts should scrutinize future ClaimClam submissions, something plaintiffs and defendants will no doubt continue to demand.