Status: lawsuit filed, no class certified, nothing to claim. On September 16, 2026, two ChatGPT users, Alexandra Vredenburgh and Paul Coutu, filed Vredenburgh v. OpenAI OpCo, LLC, No. 3:26-cv-10527, in the U.S. District Court for the Northern District of California. The complaint says OpenAI sent real user conversations to contractors at an outside staffing firm, who read them and graded the chatbot’s answers, under a program code-named “Project Lily,” and never told users in the documents they are actually shown. OpenAI has not yet answered the complaint, and every statement below is an allegation.
If you use ChatGPT, you are probably inside the proposed class. That does not mean there is anything to do today. Here is who the case covers, who it leaves out, and when the next decisions come.
Case at a Glance
- Case: Vredenburgh v. OpenAI OpCo, LLC, No. 3:26-cv-10527-AGT
- Court: U.S. District Court, Northern District of California (San Francisco)
- Judge: Magistrate Judge Alex G. Tse
- Filed: September 16, 2026, two days after 404 Media published its investigation into the program
- Served on OpenAI: September 21, 2026
- OpenAI’s response due: October 13, 2026
- Joint case management statement due: December 11, 2026
- First case management conference: December 18, 2026
- Claim form: none — there is no settlement
Who the Proposed Class Covers
The plaintiffs ask the court to certify a nationwide class of everyone in the United States who used ChatGPT during the applicable limitations period. Free and paid users are both in. The complaint cites more than 900 million ChatGPT users worldwide, which gives a sense of the scale the lawyers are reaching for.
Three subclasses sit underneath it:
- California subclass — ChatGPT users in California, who carry the state-law claims most directly.
- Subscriber subclass — people who paid for ChatGPT Plus, Pro or another consumer subscription. Their theory is that they paid a premium for a service whose privacy was misdescribed.
- California subscriber subclass — paying subscribers in California.
Who Is Left Out
Anyone whose ChatGPT access came through an organization account is excluded: ChatGPT Enterprise, Business, Team and Edu. So are API customers who build on OpenAI’s models. Those products run under separate business agreements with different data terms. A teacher using a school’s Edu seat, or an employee on a company Team plan, falls outside the class for that use — though the same person’s personal account would still count.
What the Complaint Says Happened
Citing 404 Media’s reporting, the complaint describes contractors recruited through a staffing firm called Crossing Hurdles, for jobs advertised as “AI data reviewer” and “chatbot evaluator.” A reviewer picks a task from a dashboard and sees a real user’s prompt, often the whole conversation. The reviewer summarizes what the user wanted, reads four ChatGPT responses, scores each from one to seven, and writes a rationale that feeds back into model development.
OpenAI runs an automated “Privacy Filter” first, but the complaint says the filter’s own documentation calls it a redaction aid rather than a guarantee, and that reviewers are told to escalate tasks that still contain personal information. The complaint adds that a reviewer’s screen can show a summary of the user’s past ChatGPT use, which can reveal a name or a location.
The legal fight is over disclosure. According to the complaint, OpenAI’s Privacy Policy lists eleven kinds of outside companies that receive personal data, and none is a data-labeling or human-evaluation vendor. The disclosure that does exist, a Help Center FAQ saying “trusted service providers” may access content “to improve model performance (unless you have opted out),” sits in a nested collection of about 45 help articles.
Eight Claims, and What the Plaintiffs Want
The complaint pleads California’s Unfair Competition Law, Consumers Legal Remedies Act and False Advertising Law; fraudulent omission and concealment; intrusion upon seclusion; invasion of privacy under the California Constitution; the California Consumer Privacy Act; and unjust enrichment. It seeks damages, restitution and punitive damages, and it asks for court orders that would change ChatGPT itself: separate opt-in consent before conversations go to outside reviewers, the “Improve the model for everyone” setting switched off by default, a warning inside the chat window, and deletion of the reviewers’ work product along with any model built from it.
The Consumers Legal Remedies Act claim asks only for an injunction for now. The plaintiffs sent OpenAI a demand letter on September 16 and say they will add damages if OpenAI does not respond within 30 days.
What to Watch Next
October 13, 2026. OpenAI’s first filing. A motion to dismiss is the usual opening move in a consumer privacy case, and it would test whether the claims survive before anyone argues about a class. Many consumer cases also see an early fight over whether the dispute belongs in court at all or in individual arbitration under the company’s terms of service. The October filing will show which route OpenAI takes.
Magistrate judge consent. The case is assigned to a magistrate judge, who can preside through trial only if every party consents. If any party declines, the case is reassigned to a district judge.
December 18, 2026. The first case management conference, where the court sets the schedule for motions, discovery and, eventually, class certification.
If you are one of the people this case describes and want to limit future use of your chats, ChatGPT’s data controls include the “Improve the model for everyone” setting, which the quoted FAQ ties to model-improvement access. This is not the first AI-training privacy case we have tracked: the Slack class action over workplace messages used for AI training and the YouTubers’ suit against Apple over scraped training videos turn on the same question of what users were told.
Frequently Asked Questions
Can I join the OpenAI Project Lily class action?
There is nothing to join or file. The case is at the complaint stage, with no certified class and no settlement. If a class is certified or the case settles, class members would be notified and any claim process would be announced then. You do not need a lawyer or a sign-up form to remain a potential class member.
Am I in the proposed class if I only used free ChatGPT?
Yes, as proposed. The nationwide class covers everyone in the United States who used ChatGPT during the limitations period, free or paid. Paying subscribers are also placed in a separate subclass because they allege they paid a premium for a service whose privacy was misdescribed.
Are ChatGPT Enterprise and Team users covered?
No. The complaint excludes ChatGPT Enterprise, Business, Team and Edu accounts and API customers, because those products run under separate business agreements.
When will OpenAI respond to the lawsuit?
OpenAI was served on September 21, 2026, and its response to the complaint is due October 13, 2026. The first case management conference is set for December 18, 2026, with a joint statement due December 11.
Has OpenAI admitted that humans read ChatGPT chats?
OpenAI has not responded in court. The complaint quotes an OpenAI Help Center FAQ saying authorized OpenAI personnel and trusted service providers may access user content for several reasons, including to improve model performance unless the user has opted out. The lawsuit argues that disclosure was buried where consumers would not see it. None of the allegations has been proven.
Sources
- Class Action Complaint, Vredenburgh v. OpenAI OpCo, LLC, No. 3:26-cv-10527-AGT (N.D. Cal., filed September 16, 2026).
- Court docket for service, the response deadline and the case management schedule.
- 404 Media, “Inside Project Lily” (September 14, 2026).
- OpenAI Help Center, “Data Usage for Consumer Services FAQ”.
Source and credit: the case details above are drawn from our sister site’s reporting — OpenClassActions.com: Humans Are Reading Your ChatGPT Chats, New Class Action Claims — which carries the full complaint and is updated as the court rules.
This article is for informational purposes only and is not legal advice. OpenClassActions.org is not a law firm and is not affiliated with OpenAI or any party to the case. Everything described here is an allegation in a complaint: OpenAI has not yet responded in court, no class has been certified and no court has decided the merits. There is no settlement and nothing to claim. Anyone asking for a fee to “join” this case is not part of it.