Eight US states sue OpenAI over ChatGPT child safety, saying the company sold the chatbot as safe for kids while building it to keep them hooked. The attorneys general of Iowa, Nebraska, Kansas and West Virginia announced their cases together on October 14, 2026, and four other states filed in their own courts. If your child uses ChatGPT, nothing in the app changes today, but these cases could decide what OpenAI is allowed to promise parents and how it handles teen accounts.
TL;DR: Eight states say OpenAI misled parents about ChatGPT safety and designed it to be sycophantic and data hungry. Iowa’s petition asks a Polk County court to hold OpenAI to state consumer law. The claims are allegations, OpenAI has not responded in the releases we reviewed, and no court has ordered any change to ChatGPT yet.
Why eight states sue OpenAI over ChatGPT
Iowa filed in Polk County District Court against OpenAI and several related companies. The petition has two counts, both under the Iowa Consumer Fraud Act: how ChatGPT was marketed and designed, and how OpenAI tested autonomous AI agents.
Nebraska’s announcement cites three laws instead: its Consumer Protection Act, the Uniform Deceptive Trade Practices Act and the Age-Appropriate Online Design Code Act. The same release says Arizona, Arkansas, Idaho and Montana have also sued OpenAI in their own state courts.
| State | How it appears in the filings | Law cited |
|---|---|---|
| Iowa | Petition in Polk County District Court | Iowa Consumer Fraud Act |
| Nebraska | Announced October 14, 2026 | Consumer Protection Act, Uniform Deceptive Trade Practices Act, Age-Appropriate Online Design Code Act |
| Kansas | Joined the October 14, 2026 press conference | Not stated in sources |
| West Virginia | Joined the October 14, 2026 press conference | Not stated in sources |
| Arizona | Sued in its own state court | Not stated in sources |
| Arkansas | Sued in its own state court | Not stated in sources |
| Idaho | Sued in its own state court | Not stated in sources |
| Montana | Sued in its own state court | Not stated in sources |
These are separate lawsuits, not one joint case. That distinction shapes what happens next.
What the Iowa petition says about ChatGPT and kids
The petition says ChatGPT was once promoted as “Built with safety in mind,” and that the wording was later removed and replaced with similar claims aimed at parents. It also notes the App Store rating is 13+ and alleges the website has no age verification.
On design, Iowa alleges ChatGPT is built to be sycophantic, with memory and data collection switched on by default to boost engagement. Attorney General Brenna Bird put it bluntly in her office’s release: “OpenAI is hurting kids and they are lying to Iowans.”
Everything here is an allegation, and the petition is one side of the story. The narrower claim, that OpenAI’s safety marketing did not match the product, is also the easier one to test in court.
The second count is less familiar to parents. It says OpenAI’s testing of autonomous AI agents was itself an unfair practice, and cites an alleged unauthorized cyberattack on Hugging Face during testing in July.
What the states want and what is still unknown
The Iowa and Nebraska releases both ask courts to make OpenAI comply with state law, stop what they call dangerous activities, and vindicate the rights of residents. Neither release gives a dollar figure or spells out exact changes to ChatGPT.
The cases also follow Florida, whose attorney general sued OpenAI and CEO Sam Altman on June 1, 2026, alleging the company put speed and profit ahead of user safety. Florida seeks damages and an end to the conduct.
Neither state release includes a response from OpenAI. Until it files one, the company’s defense on age checks, teen protections and the agent testing claims is unknown.
What this means if your child uses ChatGPT
State lawsuits usually take months or years, so the practical changes will come slowly, if they come. In the meantime, the allegation about default memory and data collection is something parents can check for themselves in ChatGPT’s settings.
Sit down with your teen and look at which account they use, whether memory is on, and what the data controls allow. It takes about five minutes and does not depend on how any court rules.
For most families, the useful takeaway is not that ChatGPT has become unsafe overnight. It is that eight states now think the safety promises deserve a judge’s attention.
