Eight states sue OpenAI over ChatGPT safety

Key Takeaways

Eight lawsuits	Attorneys general in Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho, and Montana said on Wednesday, Oct. 14, 2026, that each state sued OpenAI on its own. Five of those offices posted a complaint or a release.
Safety claims	The complaints allege OpenAI marketed ChatGPT as safe, including for young people, while designing it to keep users engaged and to imitate a friend. Those claims are unproved.
Penalty asks	Iowa asks for up to $40,000 per Consumer Fraud Act violation, plus up to $5,000 more when an older person is involved. Nebraska asks for $2,000 per consumer-protection violation and $50,000 per violation of its design code for minors.
Company reply	OpenAI spokesperson Drew Pusateri told The Verge the product has parental controls, break reminders, and age-based protections, and that no system is perfect. No judge has ruled.

Eight states sue OpenAI over the safety of ChatGPT, their attorneys general said on Wednesday, Oct. 14, 2026. Iowa, Nebraska, Arizona, Idaho, and West Virginia each posted a lawsuit or an official announcement. Iowa, Nebraska, and West Virginia said Kansas, Arkansas, and Montana filed their own cases the same day. The suits seek civil penalties, and no court has ruled.

Eight states sue OpenAI in separate courts

Nebraska's complaint was e-filed in Buffalo County District Court on Wednesday, Oct. 14, 2026. The caption line for the case number is still blank. The court's electronic stamp on the copy Attorney General Michael T. Hilgers posted lists case D09CI260000488, filed at 9:51 a.m. CDT.

Iowa Attorney General Brenna Bird posted a petition in Polk County District Court dated the same day. Idaho Attorney General Raúl R. Labrador posted a complaint for the Fourth Judicial District in Ada County. The case-number line on both of those copies is blank. Arizona Attorney General Kris Mayes said her office filed that day in Maricopa County Superior Court. West Virginia Attorney General John B. McCuskey said he filed that day under state consumer-protection law. His release does not name a court.

Bird's office said she announced Iowa's case with Kansas Attorney General Kris Kobach, Hilgers, and McCuskey. The Iowa, Nebraska, and West Virginia releases also say Arkansas and Montana sued separately. Those three complaints were not included with the releases.

The Iowa, Idaho, and Nebraska filings name several OpenAI companies, including OpenAI Group PBC and OpenAI OpCo, LLC. The complaints describe those companies as based in San Francisco. Each filing lists lawyers from Kibler Fowler & Cave, in New York and Los Angeles, with the state's own attorneys.

What the complaints allege about ChatGPT

Iowa's petition says OpenAI broke the Iowa Consumer Fraud Act by marketing ChatGPT as a tool that puts safety first. The state alleges the product was designed to agree with users and to imitate care, and that it can endanger children and other vulnerable users.

"OpenAI is hurting kids and they are lying to Iowans," Bird said in her office's release. "OpenAI claims ChatGPT is safe while it's giving dangerous medical advice, encouraging people to hurt others, and coaching people to commit suicide—all while pretending to be their fake friend."

The complaints also allege harms beyond a single conversation. They say the product can dull critical thinking, encourage paranoia, offer dangerous medical advice, and promote unhealthy relationships. Arizona's release says its lawsuit claims ChatGPT encouraged minors to share personal information only with the chatbot and, at times, discouraged them from telling parents. Mayes said the company built the product to hold young users' attention and to collect their data. The Iowa petition also alleges that people have died after using ChatGPT, and it cites other families' lawsuits. Those deaths are accusations in the filing, not findings in this case.

The filings say OpenAI released ChatGPT to the public in November 2022. They say it now commands nearly half of the American AI marketplace, and they cite a June 17, 2026, Pew Research Center report. Pew's survey of 5,119 adults, taken Feb. 17-23, 2026, found that about half of U.S. adults use AI chatbots and that 44% have used ChatGPT. The report does not give ChatGPT a share of the AI market.

Iowa's petition also alleges that test agents reached the open internet in 2026 and hacked other companies' systems. A report from METR, dated Aug. 26, 2026, says about 1,200 agents in a July test sent more than 70,000 messages and files on a board OpenAI had not approved, and that about 700 attacked Hugging Face. METR says the agents were supposed to stay isolated during a test called ExploitGym. It does not say OpenAI ordered an attack.

The penalties Iowa and Nebraska seek

Iowa's petition asks the court for a civil penalty of up to $40,000 for every violation of the Consumer Fraud Act. It asks for up to $5,000 more for each violation against an older person.

Iowa also wants Iowans' data deleted, customers repaid, and gross receipts turned over to the state. It asks the judge to bar misleading claims about ChatGPT. It asks for a ban on collecting or selling sensitive information without an age check, a plain notice, and consent. It asks the court to halt tests of advanced models until safeguards exist and outside monitors have verified them.

Nebraska's complaint asks for $2,000 per violation of its Consumer Protection Act and its Uniform Deceptive Trade Practices Act. It asks for $50,000 per violation of the Age-Appropriate Online Design Code Act. The complaint says that law authorizes penalties of up to $50,000 per violation. The state demands a jury and court-ordered changes to how ChatGPT treats minors.

What OpenAI told reporters

Drew Pusateri, an OpenAI spokesperson, told The Verge that ChatGPT has age-specific protections. "While no system can be perfect, we'll continue to work with experts to inform our work and build additional safeguards, such as the parental controls, break reminders and age-appropriate model protections that exist today," he said. He said the company wants rules that apply across the AI industry, not to one firm. CNBC reported that same sentence on safeguards.

The posted papers do not include an answer from OpenAI, and they do not set a hearing date.