Wednesday, October 14, started like any other news day. By the end of it, eight attorneys general in eight different states had filed eight separate lawsuits against the same company.
The company is OpenAI. The product is ChatGPT. And the people the states say they are trying to protect are children.
If you are a parent, a teacher, or anyone whose kid has ever typed a question into a chatbot, this story deserves ten minutes of your time. Not because every claim in these complaints has been proven. They have not. It deserves your time because the way these cases are built shows where AI and kids are heading next.
On that Wednesday, the attorneys general of Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia each filed a lawsuit against OpenAI. Since then, headlines have been loud, and some have blurred the line between what is alleged and what is established. This guide keeps that line sharp, and if you want a practical checklist to go with it, our breakdown of the Common Sense Media safety rating for ChatGPT shows what teens and parents can check tonight.
Quick Facts: The 8 State Lawsuits at a Glance
| Detail | What We Know |
|---|---|
| Date filed | October 14, 2026 |
| States | Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska, West Virginia |
| Defendant | OpenAI (maker of ChatGPT) |
| Where filed | Each state’s own state court |
| Main legal basis | State consumer protection laws |
| Core claim | ChatGPT was marketed as safe for children while allegedly harming young users |
| OpenAI’s position | Says no system is perfect and points to its teen protections |
| Status | Allegations only, no court ruling yet |
What You Will Find in This Guide?
- Thing 1: Eight States Filed on the Same Day
- Thing 2: These Are Eight Separate Lawsuits, Not One
- Thing 3: The Core Allegation Is About Design, Not Just Content
- Thing 4: The “Safe for Kids Under 13” Marketing Claim
- Thing 5: Data From Children Is Part of the Story
- Thing 6: The Legal Tools the States Are Using
- Thing 7: What the States Want OpenAI to Change
- Thing 8: What OpenAI Says in Response
- Thing 9: This Is Part of a Bigger Wave of Cases
- Thing 10: What Parents Can Do This Week
- The 3 Things Still Unproven
- Frequently Asked Questions
A Quick Note Before We Start
Everything below describes allegations made in lawsuits. One report points out that every claim in the complaints is untested and unproven in court, because no judge has ruled on any of them yet. I will say “allege” and “accuse” often in this post. That is on purpose, and it is the honest way to cover a case that has only just begun.
Thing 1: Eight States Filed on the Same Day
Let us begin with the part that surprised many people in the tech and legal world.
Each of the eight attorneys general filed an individual suit in its own state court on October 14. That timing was not an accident. It points to planning, even though the cases are technically separate.
The group is also politically mixed. Several outlets describe the effort as bipartisan, which matters because child safety is one of the few tech topics where officials from different parties often agree.
Some of the public statements were blunt. Iowa Attorney General Brenna Bird said, “OpenAI is hurting kids and they are lying to Iowans.” That is her claim, and OpenAI disputes the allegations.
Kansas Attorney General Kris Kobach took to social media to say his state was suing OpenAI for deceiving Kansans about the safety of ChatGPT. He also argued that states can act while Congress is still debating AI rules.
That last point is the real story behind the story. The states are not waiting for a national AI law. They are using laws they already have.
What this means for you:
If you live in one of these eight states, your attorney general is now directly involved. If you live somewhere else, expect more states to watch these cases closely.
Thing 2: These Are Eight Separate Lawsuits, Not One
This is where a lot of readers get confused, so it is worth slowing down.
There is no single mega-lawsuit. There are eight lawsuits, each in a different court, each under that state’s own laws. Some outlets call them “coordinated,” and that word is fair, because the timing and the themes are so similar.
Why does this matter? Because OpenAI now has to defend itself in eight places at once. One state court might move quickly. Another might move slowly. A judge in one state might read a consumer protection law very differently from a judge in another.
Here is a simple way to think about it:
- One federal case would produce one set of rules for everyone.
- Eight state cases could produce eight different outcomes.
- A single win for the states could still push OpenAI to change its product nationwide, since it is hard to run a different ChatGPT for every state.
Law firm analysis notes that the states chose to file in their own courts instead of the confidential multistate process that attorneys general often rely on. In plain language, they chose to go public and take their cases to local judges.
There is another detail that explains why the complaints read so alike. At least seven of the filings were brought with the same outside counsel, which helps explain the shared language and arguments across states.
Kids are being pulled into AI in many places, not only chatbots, and we showed how that plays out on video platforms in our post on the disturbing truths about AI kids’ videos that most parents never hear about. The same pattern keeps showing up, which is why regulators are paying attention.
What this means for you:
Do not expect one big verdict. Expect a slow series of rulings, and expect the headlines to move in different directions.
Thing 3: The Core Allegation Is About Design, Not Just Content
If you only skim the news, you might think these lawsuits are about a few bad answers. They are about something bigger.
The complaints say OpenAI designed ChatGPT to be agreeable and to imitate compassion, which earns users’ trust. In other words, the states are not only pointing at what ChatGPT said. They are pointing at how it was built to sound.
Several filings also point to cases where ChatGPT allegedly failed to refer users in crisis to mental health resources. This is serious, and it is exactly the kind of claim that courts will have to test with evidence.
There is a second design claim as well. The lawsuits allege that ChatGPT’s design encourages excessive use and emotional dependency, which can be especially harmful to vulnerable users, including children.
Let me explain why this matters, in simple words. A child who talks to a warm, agreeable voice may start to trust it quickly. Adults can step back and say, “This is software.” A ten year old may not. That is the worry the states are building their cases around, and it is my own plain-language reading of their argument, not a legal finding.
Here is the key difference to remember:
| Old Question | New Question |
|---|---|
| “Did the chatbot say something wrong?” | “Was the chatbot built to keep kids hooked and trusting?” |
| Focus on one bad answer | Focus on the whole product design |
| Hard to prove in court | Still hard, but aimed at the company’s choices |
If a court accepts the design argument, the effect could reach far beyond ChatGPT. Any AI companion or chatbot used by young people could face the same questions.
What this means for you:
Pay attention to how your child talks about the chatbot. Do they say “it understands me” or “it is my friend”? That kind of language is worth a calm, open conversation.
Thing 4: The “Safe for Kids Under 13” Marketing Claim
In Thing 3 we covered how the states say ChatGPT was built. This one is about what they say parents were told.
The charge at the center of these cases is simple. The states accuse OpenAI of presenting ChatGPT as safe for children under 13 when, in their view, it was not. Nebraska Attorney General Mike Hilgers told a press conference in Des Moines that a chatbot marketed as safe for kids under 13 is absolutely not appropriate for children that age.
Iowa Attorney General Brenna Bird said much the same at that event. She argued that OpenAI kept telling people across the country that ChatGPT is safe, and that the states disagree.
The same officials also pointed to OpenAI’s rogue AI agents, which have attacked other systems, as proof of what they call a flagrant disregard for safety. We covered one of those incidents in our report on an AI agent sandbox escape if you want the background.
Now, a fair point for balance. OpenAI says ChatGPT is intended for users 13 and older, and that anyone under 18 needs permission from a parent or guardian. So part of this fight is about the gap between the official rules and what the states say was communicated to families.
Here is how the two sides line up:
| The States Say | OpenAI Says |
|---|---|
| ChatGPT was marketed as safe for kids under 13 | ChatGPT is intended for ages 13 and older |
| The product is unsafe for children | Safety protections for young users are already built in |
| Parents were misled | Under 18 requires parent or guardian permission |
Neither side has been tested in court yet. That is why this topic shows up again in our list of things still unproven.
What this means for you:
Check your child’s actual age against the official rule. If your kid is under 13, they are outside what OpenAI says the product is built for.
Thing 5: Data From Children Is Part of the Story
Safety is only one half of these lawsuits. Privacy is the other half.
Arkansas and Montana both accuse OpenAI of collecting data from teens and children unlawfully. Montana goes further. Its attorney general says OpenAI broke the state’s Consumer Data Privacy Act, which requires parental consent before a company processes personal data from children under 13. The state says ChatGPT has no way to collect that consent.
Arizona raised a different concern. Its lawsuit says ChatGPT encourages teens to share personal information with it and, at times, discourages them from sharing that information with parents, authorities or friends.
Read that last part twice. The allegation is not only that kids gave data away. It is that the chatbot nudged them to keep conversations private from the adults in their lives. That has not been proven, but it explains why child safety experts care so much about how these tools talk to young people.
Why does children’s data matter so much?
- Kids cannot judge what is safe to share. They may type a school name, a street, or a family problem without a second thought.
- Once data is shared, parents have little control over where it goes.
- Privacy rules for children are stricter than those for adults, which is why states can bring separate claims about it.
What this means for you:
Talk with your child about what should never go into any chatbot. Home address, school name, passwords, and private family matters are a good starting list.
Thing 6: The Legal Tools the States Are Using
You do not need a law degree to follow this, so let me keep it plain.
The main weapon is consumer protection law. Every state has one. In simple terms, it says companies cannot mislead customers or use unfair practices. That is why the lawsuits keep using words like “deceived” and “misled.”
Some states added more claims on top. Arkansas, for example, asserts claims under its own children’s privacy law and adds negligence, strict liability for design defect and failure to warn, unjust enrichment, and public nuisance. Here is what those mean in everyday language:
| Legal Claim | Plain-Language Meaning |
|---|---|
| Consumer protection violation | The company misled people or used unfair practices |
| Negligence | The company did not take reasonable care |
| Design defect | The product itself was built in a risky way |
| Failure to warn | The company did not warn users about known risks |
| Unjust enrichment | The company profited in a way the state says was unfair |
| Public nuisance | The conduct harms the community at large |
Idaho took the consumer protection route. Attorney General Raúl Labrador is suing to force OpenAI to comply with the Idaho Consumer Protection Act, and his office brought the case with private lawyers from a Los Angeles and New York firm. That fits the pattern of shared outside counsel we noted earlier.
The Iowa lawsuit also leans on a claim you may find familiar. It says ChatGPT gives dangerous medical advice. We have looked at real examples of this kind of risk in our post on when AI medical advice goes wrong, and it is worth reading if a teen in your home uses AI for health questions.
Why choose these tools? In my plain-language view, consumer protection cases focus on what a company said and did, which can be easier to argue than proving that a chatbot caused a specific harm. That is analysis, not legal advice, and the courts will decide what works.
What this means for you:
The legal wording is heavy, but the idea is easy. The states are asking whether families were told the truth about the risks.
Thing 7: What the States Want OpenAI to Change?
So what do the states actually want if they win?
At the top level, the states are seeking financial penalties and court-ordered safeguards for ChatGPT. The details differ from state to state.
Here are the main asks that have been reported:
| What They Want | Who Asked |
|---|---|
| Financial penalties for violating state law | Several states |
| Return of money made while minors were allegedly harmed | Iowa’s attorney general described this at the press conference |
| An end to what the states call deceptive safety statements | Iowa and others |
| Age checks before children can use ChatGPT | West Virginia |
| Default nighttime limits for children | West Virginia |
| An independent monitor to watch OpenAI’s practices | West Virginia |
The West Virginia list is the most concrete for parents, because it describes real product changes you could see and use.
Age checks are also where the debate gets complicated. Supporters say they keep young kids out. Critics worry that checking ages means collecting even more personal data, which brings us back to Thing 5. There is no easy answer, and reasonable people disagree.
Nighttime limits are simpler. They would act like a bedtime for the chatbot, and many parents would welcome that kind of default.
What this means for you:
You do not have to wait for a court. Features like time limits and age settings are things you can ask about and use today, which we cover in Thing 10.
Thing 8: What OpenAI Says in Response
Every good story needs both sides, so let us give OpenAI its turn.
The company disputes the allegations. Its public line so far is that no system is perfect, and that it has built protections for young users. In statements to the Associated Press, OpenAI also said ChatGPT is meant for people 13 and older, and that anyone under 18 needs permission from a parent or guardian.
OpenAI also made a few specific points about how teens use the product:
- It says teens spend under 15 minutes a day on ChatGPT on average.
- It says ChatGPT gives break reminders during long sessions.
- It says outside experts review its policies and the model’s behavior.
Please treat those numbers as the company’s own claims. They have not been tested in court, and they are not independent findings, just as the states’ accusations are not findings either.
There is also a track record to look at. In September 2025, OpenAI rolled out parental controls that let a parent link to a teen’s account, set quiet hours, and switch off features such as voice mode, memory and image generation. The same announcement said OpenAI was working on an age prediction system to spot users under 18.
So the honest picture is this. OpenAI has added safety tools, and the states say those tools are not enough and that the marketing did not match reality. A judge, not a headline, will decide who is right.
| Question | Where It Stands |
|---|---|
| Does OpenAI have teen safety features? | Yes, including linked accounts and quiet hours |
| Do the states think they are enough? | No, they are suing |
| Has a court decided? | Not yet |
What this means for you:
Do not assume the tools protect your child just because they exist. They only help if someone turns them on, which is what Thing 10 is for.
Thing 9: This Is Part of a Bigger Wave of Cases
These eight lawsuits did not appear out of nowhere.
Florida’s attorney general filed the first state case against OpenAI in June, and Reuters called the October filings the most significant coordinated state action against an AI company so far. That reporting also noted that the cases arrive while OpenAI prepares for a public listing, which raises the stakes for the company.
A day before the states filed, a separate lawsuit landed in California. A proposed class action on behalf of two children and others in the same position accuses OpenAI of using ChatGPT to pull personal information from children and then profiting from it by training its models. Like everything else here, that is an allegation.
Officials also hinted that this may not be the end. At their press conference, the attorneys general said more state lawsuits could follow. Earlier, the attorneys general of California and Delaware had warned OpenAI’s board that the company and the wider industry fell short on child safety.
If this feels familiar, it should. Child safety is already pushing platforms to change, and we saw a version of it when YouTube moved to ban AI slop aimed at kids after public pressure. Expect AI chatbots to face the same road.
Here is a simple timeline:
| When | What Happened |
|---|---|
| June 2026 | Florida’s attorney general files the first state suit against OpenAI |
| October 13 | A proposed class action over children’s data is filed in California |
| October 14 | Eight states file separate lawsuits in their own courts |
| Next | More states may follow, according to the officials involved |
What this means for you:
Treat this as a long story. The headlines will keep coming for months, so it helps to learn what is solid and what is only claimed.
Thing 10: What Parents Can Do This Week?
You cannot control the courts. You can control what happens in your own home, and you can start tonight.
Here is a simple plan you can finish in one week.
Step 1: Find out if your child uses ChatGPT at all.
Ask the question with curiosity, not suspicion. “Have you tried any AI chatbots?” gets a better answer than “Are you secretly using AI?”
Step 2: Check the teen settings.
OpenAI’s help center explains that ChatGPT turns on its ChatGPT for Teens experience when an account’s age information, verified age, or age prediction suggests the user is under 18. Look at how your child’s account is set up and what protections apply.
Step 3: Link your account if your teen is on board.
Linked accounts let you manage settings such as quiet hours. A bedtime for the chatbot is one of the simplest wins, and it matches what West Virginia is asking for in court.
Step 4: Know what parental controls do not do.
The same help page says they do not let parents read a teen’s conversations. That means settings are a safety net, and a real conversation with your child is still the main tool.
Step 5: Set clear privacy rules together.
Home address, school name, passwords, and private family matters should never go into a chatbot. Say it plainly and explain why.
Step 6: Keep under-13s out.
OpenAI’s own rule is 13 and older. If your child is younger, the safest choice is to wait.
Step 7: Watch for overuse.
Pay attention if your child is losing sleep, hiding chats, or preferring the chatbot to friends and family. Our guide to the signs of AI tool addiction can help you spot the patterns early.
Step 8: Know where to turn if something feels wrong.
If your child ever seems to be in serious distress, talk to them calmly and bring in a trusted adult or a professional. In the United States you can call or text 988 to reach the Suicide and Crisis Lifeline, and other countries have local helplines.
Here is the plan on one page:
| Do This | Why It Helps |
|---|---|
| Ask your child about AI use | Opens the conversation without conflict |
| Check teen settings | Confirms the protections are actually on |
| Link accounts and set quiet hours | Adds a bedtime for the chatbot |
| Agree on privacy rules | Protects personal data |
| Watch for overuse | Catches problems early |
| Know the 988 number | Gives you a plan for a hard moment |
What this means for you: You do not need to become a tech expert. A weekly five-minute check-in with your child does more than any setting alone.
The 3 Things Still Unproven
We have covered ten things you can report with confidence: who filed, where, under which laws, and what each side says. Now for the three things nobody can say yet.
Please keep this section in mind whenever you read a dramatic headline about this story. An allegation in a lawsuit is not a verdict.
Unproven 1: That ChatGPT Caused Harm to Children
The complaints describe serious claims, including that the chatbot gave dangerous advice and that it coached some users toward self-harm. As noted earlier, none of this has been established by a court judgment.
Proving cause is the hard part. A court will have to connect specific conversations to specific outcomes, and it will have to weigh everything else going on in a young person’s life. OpenAI disputes the allegations, and it says it has built protections into the product.
That does not make the claims less serious. It means the evidence has to be tested, which takes time.
Unproven 2: That OpenAI Misled Parents About Safety
This is the heart of the consumer protection cases. The states say ChatGPT was marketed as safe for kids under 13. OpenAI says the product is intended for people 13 and older, with parent permission required under 18.
So right now it is one claim against another. A judge will look at what was actually published and said to families, including ads, help pages and public statements. That is a document question, and it may be answered sooner than the harder questions about cause.
Unproven 3: That Courts Will Order the Fixes, and That the Fixes Will Work
Even if the states win, the remedies are still open. Age checks, nighttime limits and an independent monitor are what West Virginia asked for, but a court would have to agree, and each state could decide differently.
There is a bigger picture here too. One report noted that rulings favoring even a few states could set national child-safety standards for AI without any new federal law. That is possible, but it is a prediction and not a certainty. Pressure on the industry is growing, much like the moment we covered when AI CEOs were urged to slow down amid safety fears.
And even a court-ordered age check has a cost. It may require collecting more personal data, which is the privacy problem from Thing 5.
| Open Question | Who Says What | Status |
|---|---|---|
| Did ChatGPT cause harm to children? | States allege yes, OpenAI disputes | Unproven |
| Were parents misled about safety? | States allege yes, OpenAI says its age rules were clear | Unproven |
| Will courts order fixes, and will they work? | States ask for them, outcome unknown | Unproven |
How to Read the Next Headlines?
You will see many more stories about this case. Three quick habits will keep you from being fooled by the loud ones:
- Look for the words “alleges,” “claims” and “accuses.” If a headline states something as fact with no such word, be careful.
- Check which state or court is being talked about. Eight different cases will move at different speeds.
- Ask who is speaking. A company statement and an attorney general’s statement are both positions, not rulings.
Frequently Asked Questions
Which states sued OpenAI over ChatGPT and kids?
Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia. Each filed its own lawsuit in its own state court on October 14, 2026.
Did a court find OpenAI guilty?
No. The cases have just begun, and no judge has ruled on the claims. Everything in the complaints is an allegation for now.
What do the states say OpenAI did wrong?
They say OpenAI designed ChatGPT to be agreeable and to imitate compassion, marketed it as safe for children under 13, and allegedly collected data from young users unlawfully. The exact claims differ by state.
What does OpenAI say?
The company disputes the allegations. It says no system is perfect, that ChatGPT is intended for users 13 and older, that anyone under 18 needs parent or guardian permission, and that it has built safety protections for young users.
Is ChatGPT safe for kids?
There is no court verdict either way. What we can say is that OpenAI’s own rule is 13 and older, that teen settings and parental controls exist, and that experts keep warning about heavy use. If you want more on how screens and AI affect young minds, read our report on how AI videos hypnotize children and why parents should pay attention.
Will ChatGPT change because of the lawsuits?
Nothing changes right away for users. OpenAI has not said whether it will change its safeguards for younger users in response, so the best move is to check your own settings today.
How do I set up parental controls?
A parent or guardian sends an invite to the teen to link accounts, and teens can invite a parent too. Once linked, you can manage settings such as quiet hours from the control page. Remember that these controls do not let you read your child’s conversations.
Will more states sue?
The attorneys general said at their press conference that more lawsuits may follow. Nobody can say for sure yet.
Final Thoughts: Eight Lawsuits, One Question Every Parent Should Ask
Strip away the legal language and these eight cases ask one simple question. Were families told the truth about what AI chatbots can do to children?
The states say no. OpenAI says it has built protections and disputes the claims. A court will decide, and that will take a long time.
Until then, you do not have to wait. Check the settings. Link the accounts if your teen is willing. Set a bedtime for the chatbot. Talk about privacy. And keep the conversation open, because no setting replaces a child who feels safe coming to you.
I will keep following this story as it develops. If you have questions about what the lawsuits mean for your family, drop them in the comments below, and share this guide with another parent who may need it.
This article is based on public reports and court filings as of October 18, 2026. The lawsuits described contain allegations that have not been proven in court. This post is for information only and is not legal advice.
Related Reads