Meta NameTag and AI training facial recognition lawsuit

Meta Platforms faces a proposed class action alleging that it took faces from photos people posted to Facebook and Instagram and used them to build two kinds of AI without asking. One is NameTag, a face recognition feature designed for Ray-Ban and Oakley Meta glasses and the Meta AI app. The other is a family of image generating models, including Emu and Muse Image.

The case is Alvarez v. Meta Platforms, Inc., No. 1:26-cv-10773, filed September 4, 2026, in the U.S. District Court for the Northern District of Illinois and assigned to Judge April M. Perry. It is a lawsuit, not a settlement. There is no claim form and no money available today.

The plaintiffs are two parents suing for themselves and their children. Francisco Alvarez of Illinois has used Facebook for about 15 years and Instagram for about 10, according to the complaint. His son, identified as F.M.A., appeared in a public Instagram profile photo when he was about three. Jeremy Wahl of California says he posted photos of his daughter, I.W., now 10, throughout her life on accounts that were public for years.

Their lawyers are Wexler Boley & Elgersma LLP and Stephan Zouras, LLC of Chicago, with Gustafson Gluek PLLC of Minneapolis and Taus, Cebulash & Landau, LLP also listed. The defendant is Meta Platforms, Inc., which owns Facebook, Instagram and the Meta AI app and sells smart glasses with Ray-Ban and Oakley.

According to the complaint, WIRED reported in June 2026 that Meta had placed face recognition code, called NameTag, inside the Meta AI companion app, which the complaint says has been downloaded more than 50 million times. The complaint quotes WIRED's description that, once active, NameTag would turn faces captured by the glasses into faceprints and compare them to faceprints stored on the user's phone. Meta told WIRED at the time that "Nothing has shipped to consumers and no final decision has been made." WIRED later reported that Meta removed the code the day after the June story.

The complaint also describes Meta's image AI work. It says Meta has stated it used 1.1 billion image and text pairs to train its Emu model. It says that when Meta launched Muse Image in July 2026, an Instagram feature let people generate images of public account holders, and that Meta pulled the feature three days later, saying, "We've heard the feedback that this feature missed the mark." The plaintiffs allege the same capability stayed available through the Meta AI app and website.

The plaintiffs allege that each time Meta trained a face recognition or image model on a photo, it created and stored biometric identifiers, such as face geometry, without the written notice and consent Illinois law requires. They also allege that using people's likenesses to build commercial AI products violated California publicity and privacy rights. The complaint lists five claims:

• Violation of section 15(a) of the Illinois Biometric Information Privacy Act, which covers retention policies.
• Violation of section 15(b) of the same law, which covers notice and written consent.
• California common law right of publicity.
• Commercial misappropriation under California Civil Code section 3344.
• Invasion of privacy under the California Constitution.

The complaint cites Meta's past privacy penalties, including a $650 million Illinois biometric settlement and a $1.4 billion Texas biometric settlement, as context.

Meta disputes the claims. In a statement reported by WIRED on September 11, 2026, Meta said, "This lawsuit is without merit and misrepresents our work," adding that it has been transparent about how it uses people's information to build AI and that "we are not building a universal face database." Meta has not yet filed a response to the complaint on the docket we reviewed.

The court has decided only scheduling matters. The docket shows that Meta asked for more time to respond to the complaint on September 25, 2026, and the court granted that request on September 28. In mid-September, several out-of-state lawyers filed to appear in the case. On October 23, a party asked permission to file a brief longer than the usual page limit; the docket entry does not describe the brief.

The court has not ruled on whether Meta collected biometric data, whether any law was broken or whether the case can go forward as a class action.

The complaint proposes three classes, each limited to the period from September 4, 2021, to the present:

• Illinois class: all persons in Illinois whose image was uploaded to a Facebook account, an Instagram account or a Meta generative AI model during the inference stage.
• National class: all persons in the United States whose image was uploaded the same ways.
• California class: all persons in California whose image was uploaded the same ways.

Meta, entities it controls, and its legal representatives, officers and directors are excluded. The definitions turn on whose image was uploaded, so they could reach people who appear in photos posted by others. The complaint estimates the national class in the millions. A judge must approve any class, and definitions often narrow.

1. Write down the Facebook and Instagram accounts you have used, roughly when you opened them, and whether they were public.
2. Parents can note accounts where photos of their children were posted.
3. Review Meta's privacy settings and the visibility of your profile photo and older albums.
4. Watch for court-approved notice. Anything official will name the case and the court.

There is no payment now. The complaint asks the court to award the Illinois class the greater of actual damages or $1,000 for each negligent violation and $5,000 for each intentional or reckless violation of the Illinois Biometric Information Privacy Act. For the national and California classes, it seeks statutory, compensatory, punitive and other damages under California law, including Civil Code section 3344. It also asks for a permanent injunction and attorneys' fees. These are requests, not awards. Any recovery would depend on a court judgment or settlement.

• September 4, 2021: start of the proposed class period.
• June 2026: WIRED reports NameTag code in the Meta AI app.
• July 2026: Meta launches Muse Image and withdraws the Instagram feature three days later, according to the complaint.
• September 4, 2026: complaint filed in Chicago.
• September 28, 2026: court grants Meta more time to respond.

There are no claim, opt-out or objection deadlines. Claims under biometric and privacy laws carry time limits, so anyone weighing a separate case should talk to a lawyer.

• Biometric identifier: under Illinois law, includes a scan of face geometry. A faceprint is a numeric map of a face.
• Inference stage: when a finished AI model processes a new input, such as a photo a user submits, to produce an output.
• Right of publicity: a person's right to control commercial use of their name, image or likeness.
• Proposed class: a group the plaintiffs ask to represent. It has no legal effect until a judge certifies it.

Meta is expected to answer or move to dismiss. If the case survives, the parties would gather evidence before the plaintiffs ask the judge to certify a class. That process usually takes a year or more.

This article is based on the complaint, the public docket, Meta's statements reported by WIRED, and the text of the Illinois and California laws cited. Descriptions of how NameTag and Meta's AI models work come from the complaint and WIRED's reporting, not from findings by a court. We did not independently test any Meta product. The date the court set for Meta's response was not visible on the docket we reviewed.

Class Action Pulse is not a law firm or settlement administrator and does not guarantee eligibility or payment. This article is general information, not legal advice.