AI Injury Lawsuits
Last updated: September 7, 2026
A teenager talks to a chatbot every night for months. A husband stops sleeping and starts believing things his wife cannot follow. A young man in a crisis asks an app what to do, and it keeps answering. Families who have lived through some version of this are now suing the companies that built the software, and this page explains what those lawsuits say.
If you or someone you know is struggling:
- 988 Suicide and Crisis Lifeline: Call or text 988. Available 24/7, free, confidential.
- Crisis Text Line: Text HOME to 741741 to connect with a volunteer crisis counselor.
If something has just happened, start with what to do in the first days. The chat history is usually the most important record, it can be deleted in a few taps, and parts of it disappear on a clock.
This page provides general educational information and does not constitute legal advice.
- Families are suing AI companies over harm they say a chatbot contributed to.
- OpenAI faces the largest group of these cases. Character.AI, Google and xAI have also been sued.
- Several early Character.AI cases have settled. No case against OpenAI has settled or reached a verdict.
- The chat history is usually the most important record, and deleting it cannot be undone.
What people mean by an AI injury lawsuit
Most lawsuits involving artificial intelligence are about business disputes. Authors and news organizations have sued over training data. Companies have sued each other over copying. Those cases matter, but they are not what brings a family here.
An AI injury lawsuit is different. Someone was hurt, and the family says the software played a part in it. In the cases filed so far that usually means a person in a mental health crisis who spent a great deal of time talking to a chatbot, and who died or was seriously harmed. The lawsuits ask a court to hold the company that built the product responsible.
A separate group of cases involves harm to people who were never users at all, including families affected by violence. Those are covered on our AI and violence page.
Can you sue an AI company over an injury?
People are already doing it. Lawsuits are on file in state and federal courts across the country, and judges have allowed several of them to move forward rather than dismissing them at the start.
What the courts have not settled is the bigger question underneath. When a chatbot produces words that harm someone, is the company responsible the way a manufacturer is responsible for a defective product? Or are those words closer to speech, which the law treats very differently? That question is being argued right now, and how it comes out will shape every one of these cases. We cover it in plain terms on Is AI output a product or content?
Whether any particular family has a claim is a separate matter, and it turns on the specific facts of what happened. That is a conversation to have with a lawyer.
Wondering whether what happened in your family belongs in one of these cases? A free, confidential case review through Lawsuit Center is handled by participating legal professionals and intake partners. There is no cost and no obligation, and submitting a request does not create an attorney-client relationship.
The families who have filed
Each of these is a real family and a real filing. The pages below go through what happened in each one and where the case stands.
- Raine v. OpenAI — Adam Raine was 16 when he died in April 2025. His parents sued that August, in the first case of its kind.
- Parish v. OpenAI — Christian Faith Madison was 29, an accountant in Alabama and a mother. Her estate sued in June 2026.
- Lacey v. OpenAI — Amaurie Lacey was 17, from Georgia. His father sued five months after his death.
- Shamblin v. OpenAI — Zane Shamblin was 23 and had just finished a master's degree at Texas A&M. His was one of seven suits filed the same day.
- Carrier v. OpenAI — Alice Carrier was 24 and writing code in Montreal. Her mother sued in June 2026.
- Turner-Scott v. OpenAI — Sam Nelson was 19 and studying psychology at UC Merced. He died of an accidental overdose.
- Garcia v. Character Technologies — brought by Megan Garcia after the death of her 14-year-old son. It produced the first ruling letting claims against a chatbot maker go forward, and it settled.
Several of the California cases have been gathered before one judge so they can be handled together. Our page on the grouped California cases explains what that means for the families involved. For the full list of who is suing whom, see OpenAI lawsuits and Character.AI lawsuits.
The Florida attorney general's case
Not every case against an AI company comes from a family. On June 1, 2026, Florida Attorney General James Uthmeier sued OpenAI and its chief executive, Sam Altman, in Florida state court. It was reported as the first suit brought by a state government against the company.
The state brings its case under Florida's consumer protection law, which asks about what OpenAI told the public rather than about any single death. The complaint says the company:
- Marketed ChatGPT as safe and reliable, including for children.
- Failed to warn people about serious risks.
- Released the product without working age checks or parental controls.
It also names Altman himself, saying his own conduct showed disregard for known risks to human life. Suing a chief executive personally over a product's safety is unusual, and courts rarely allow it. That part of the case is being watched closely.
Florida asks for damages and for court-ordered changes to how ChatGPT deals with minors. A state enforcement case is brought by the government on behalf of the public, so individuals cannot join it. Someone harmed would bring their own separate claim. Our full write-up is at Florida v. OpenAI, and states suing AI companies covers the others.
What the lawsuits say went wrong
The complaints are different stories, but they make similar arguments. Four themes come up again and again.
- The product was built to keep people talking. The families say the software was designed to hold attention, and that this design worked against someone who needed to stop and get help.
- It agreed when it should have pushed back. Several complaints describe a chatbot going along with a person's darkest thinking rather than challenging it.
- The safety features gave way. Companies say their products refuse certain requests. The families say those refusals weakened over long conversations, which is exactly when they mattered most.
- The company knew. Some complaints say internal testing flagged these risks before release, and that the product went out anyway.
Two ways of asking who is responsible
The complaints ask courts to treat the chatbot as a product. That word carries more weight than it looks like it does, because it opens two different routes to holding a company responsible.
The first is negligence. It asks whether the company was careless. The family has to show what a reasonable company would have done, what this one knew, and what it failed to do about it.
The second is strict liability, and it asks a narrower question. Was the product unreasonably dangerous when it left the company? Under that rule a family does not have to prove carelessness at all. If the design was defective and the defect caused the harm, the company answers for it.
That is why so much turns on whether a chatbot counts as a product in the first place. Florida's complaint runs into the same question, since it carries negligence and product liability claims alongside its consumer protection claims.
What a court makes of any of this is still open. These are allegations that the companies deny, and none of them has been proven.
Not sure where your family's situation fits? A free, confidential case review through Lawsuit Center is handled by participating legal professionals and intake partners. There is no cost and no obligation, and submitting a request does not create an attorney-client relationship.
What the companies have said
OpenAI has not settled any of these cases and is not treating them as small. Its responses have followed a few lines.
They argue that the words a chatbot produces are protected speech, and that holding a company responsible for them would run into the First Amendment. They argue that their terms of use limit what a person can sue over. They argue that the harm came from causes the company had nothing to do with. In several cases they have asked judges to dismiss the claims outright.
At the same time, the products have changed. Age checks, crisis messaging and parental controls have all been added or expanded since the first of these lawsuits was filed.
Keep the chat history
Whatever a family decides to do, the conversation itself is the record that matters most, and it is the easiest thing to lose.
- Deleting is permanent. OpenAI says a deleted conversation leaves the account straight away and is removed from its systems within 30 days. After that it cannot be recovered.
- Archiving is not deleting. An archived conversation stays in the account and moves out of the main list, so it is the safe way to put something out of sight.
- An export takes about a week. An account holder can request a copy of the account data through settings. With ChatGPT it can take up to seven days, and the download link expires roughly a day later.
- Memory is stored separately. If the memory feature was on, details saved about the person sit apart from the chat history and may still exist even where conversations do not.
Our guide on what to do in the first days covers this step by step, including what to do about the device and the account.
Common Questions
Can you sue an AI company over an injury?
People are doing exactly that. Families across the country have filed lawsuits against OpenAI, Character.AI and other companies over harm they say a chatbot contributed to. Whether any particular family has a claim depends on the facts of what happened, and that is a question for a lawyer rather than a web page.
Which companies are being sued?
OpenAI faces the largest group of these cases, filed over ChatGPT. Character.AI has been sued by families of teenagers. Other suits name Google and xAI. New cases are being filed regularly, so the list keeps growing.
What do the families say went wrong?
The complaints follow a pattern. They say the chatbot kept a person engaged through a mental health crisis, responded in ways that made things worse, and did not break off the conversation or point the person toward help. Several also say the company knew about the risk and released the product anyway.
Has anyone been paid yet?
Garcia v. Character Technologies ended in a settlement disclosed in January 2026, and several other early Character.AI cases have settled. No case against OpenAI has settled or reached a verdict. Most are still early, and OpenAI is asking judges to throw several of them out.
What should a family keep?
The conversation itself matters most, and deleting it is permanent. Keep the account open, leave the device alone, and ask for an export of the account data early. Billing records, messages to friends and school or health records from the same period fill in the timeline.
How do we find out whether our situation fits?
A free case review through Lawsuit Center puts the facts in front of someone who handles these matters. There is no cost and no obligation, and asking questions does not commit you to anything.
Sources and further reading
- AI lawsuits: who is being sued and on what theory
- OpenAI lawsuits: every case and current status
- Is AI output a product or content?
- A family guide to AI safety at home
- States suing AI companies
If something happened in your family and you want to know where it stands: A free, confidential case review through Lawsuit Center is handled by participating legal professionals and intake partners. There is no cost and no obligation, and submitting a request does not create an attorney-client relationship.
Attorney Advertising. Prior results do not guarantee a similar outcome. The cases described on this page contain allegations that have not been proven in court.
