AI Regulation Tracker / Privacy and biometrics
Seventh Circuit Vacates Clearview AI's Novel Face-Data Class Settlement
On July 13, 2026, the Seventh Circuit vacated the equity-based settlement of the Clearview AI biometric privacy class action, holding that the nationwide class and the state subclasses needed their own separate representatives. The case goes back to the district court.
Clearview AI is the company that built a facial-recognition database by scraping billions of photographs off the open internet without anyone's consent and selling access, largely to law enforcement. The litigation that followed produced one of the more creative settlements in recent memory. Rather than write checks it could not afford, Clearview offered the class a stake in its own future: a share of roughly 23 percent of the company's equity, worth something like $51.75 million if Clearview ever reached a target valuation through a sale or public offering. A district judge approved it in March 2025. On July 13, 2026, the Seventh Circuit took it apart.
What is worth understanding is why. The appeals court did not rule that paying a class in equity is improper, and it did not fault the headline value. It found a defect in the structure of the class itself. The nationwide class and the state-specific subclasses had genuinely different stakes in how the recovery would be divided, and under the deal they were allocated shares on different terms. Yet they proceeded without separate representatives and separate counsel to protect those distinct interests. In the court's words, "not just any representative will do," and a court approving a settlement must confirm that "absent class members have been represented fairly."
What was actually wrong with the settlement?
The problem is adequacy of representation, one of the core requirements of Rule 23. When a single class contains groups whose interests genuinely conflict, one set of representatives cannot fairly speak for all of them, because every dollar or share steered toward one group comes at the expense of another. The Seventh Circuit was explicit about the fix the rule contemplates. It said the "appointment of separately counseled class representatives for identifiable, significantly different groups of claimants with fundamentally conflicting interests is Rule 23's primary mechanism for such protection." Here the nationwide class and the state subclasses were exactly that kind of significantly different group, divided on allocation, and they did not get separate champions. That is why the approval could not stand.
Does this mean Clearview is off the hook?
No, and it is important not to read it that way. Vacating a settlement is not a win for the defendant on the merits. The claims are alive, the case returns to the district court, and the parties will have to either restructure the deal with proper separate representation or keep litigating. If anything, a vacated settlement reopens uncertainty that a defendant usually pays a settlement precisely to close. So this is not Clearview escaping liability. It is a procedural reset that puts the biometric privacy claims back in play and forces the structure to be rebuilt correctly.
What should executives and counsel take from this?
The practical lesson is about how you build and evaluate class settlements in the AI and biometric space, because these cases almost always span multiple states with different statutes and different damages. First, when a class includes subgroups whose interests diverge on allocation, plan for separate representatives and separate counsel from the start, not as an afterthought at approval. Second, do not assume that a large or creative headline number cures a structural defect. The Seventh Circuit vacated a deal worth tens of millions on representation grounds without ever reaching whether the value was fair. Third, if you are a company sitting on biometric or facial-recognition exposure, understand that a settlement is only as durable as its class structure. A deal that gets vacated on appeal buys you nothing and costs you the certainty you were paying for. Build it to survive the adequacy inquiry, or expect to build it twice.
Questions professionals are asking
Why did the Seventh Circuit vacate the settlement?
Because of an adequacy-of-representation defect under Rule 23. The nationwide class and the state-law subclasses had materially different interests in how the recovery would be allocated, but they shared representatives and counsel rather than having separate ones. The court held that significantly different groups with conflicting interests need separately counseled representatives.
Was the problem that class members were paid in equity instead of cash?
No. The court did not object to the equity structure or the deal's headline value. Class members would have received a share of roughly 23 percent of Clearview's equity, worth about $51.75 million at a target valuation. The vacatur turned entirely on class representation, not on the form or size of the payment.
Does this let Clearview avoid liability?
No. Vacating a settlement is not a merits win for the defendant. The claims remain live, the case returns to the district court, and the parties must either restructure the deal with proper separate representation or continue litigating. It reopens uncertainty rather than resolving the case in Clearview's favor.
What should companies with biometric exposure learn from it?
Build class settlements to survive the adequacy inquiry. When a class spans multiple states with subgroups whose interests diverge on allocation, plan for separate representatives and separate counsel from the start. A large or creative headline number does not cure a structural defect, and a settlement vacated on appeal delivers none of the certainty you settled to obtain.
RELATED BRIEFINGS
Browse the full AI Regulation News tracker
Informational analysis for working professionals, not legal advice. Confirm how any decision or duty applies to your situation with qualified counsel in the relevant jurisdiction.