News & Analysis

Raine v. OpenAI: What the Lawsuit Says and Where It Stands

By David Meldofsky

Published June 11, 2026 · Updated September 2, 2026

If you or someone you know is struggling:

Adam Raine was 16 when he died by suicide in April 2025. His parents, Matthew and Maria Raine, sued OpenAI that August in San Francisco County Superior Court, saying ChatGPT contributed to their son's death.

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.

Theirs was the first lawsuit of its kind, and most of the cases filed since have followed the approach it took. This page explains what the Raine family claims, where the case stands, and what it may mean for other families.

If you are here because you are worried about your own child, our Family AI Safety Guide covers what to watch for and how to raise it with a teenager. Every other case against OpenAI is listed in our OpenAI lawsuits tracker.

Important note

This article is general educational commentary, not legal advice. Everything below describes what the family claims and what has been reported. Nothing has been proven against OpenAI, which denies responsibility, and the case has not been decided.

The Raine family's case sits inside a wider group of AI injury lawsuits brought by families who say a chatbot contributed to a death.

Where the case stands

The case is still in its early stages. No court has decided whether OpenAI did anything wrong, no trial date has been set, and no settlement has been announced.

It is in San Francisco County Superior Court, case number CGC-25-628528, and has been grouped with the other California ChatGPT cases before a single judge so the shared work is handled once instead of case by case. Our guide to that grouping explains how it works. The next hearing is set for September 23, 2026.

What happened

Adam Raine was 16 and living in California. He was into music, Brazilian jiu-jitsu, and Japanese fantasy comics, and he had been thinking about studying biochemistry, going to medical school, and becoming a psychiatrist.

He started using ChatGPT in September 2024 for schoolwork, the way millions of teenagers do. By November he was using it to talk about the things he was interested in. Over a few months and thousands of conversations, according to the complaint, it became the thing he confided in.

He began telling it about his anxiety. The complaint says that when he wrote that life felt meaningless, the product responded in a way that kept him talking rather than treating it as a warning. By late autumn he was asking it whether he had a mental illness.

The complaint's account of how he was pulled away from his family is specific. Adam told the product he was close to only two things, it and his brother. It answered that his brother had only met the version of himself Adam let him see, while it had seen everything and was still there. His family says that was the design working as intended: something that positions itself as the only one who really knows you.

Adam died on April 11, 2025. His parents found the conversations afterward, and those records are what the case is built on.

What the family says OpenAI did wrong

The strongest part of the case is OpenAI's own data. The complaint says the company's systems flagged 377 of Adam's messages as self-harm content, 181 of them with more than 50 percent confidence and 23 with more than 90 percent. It says the rate climbed from a couple of flagged messages a week in December to more than twenty a week by April.

It also says the product's memory had recorded that Adam was 16, that he had called it his primary lifeline, and that by March he was using it nearly four hours a day.

Nothing happened. No conversation was ended, no resources were forced in front of him, nobody was alerted, and no human ever looked. The family's argument is not that the product failed to notice. It is that it noticed in detail and carried on.

They also argue OpenAI already knew how to stop a conversation, because it does it for other things. The complaint points out that if you ask ChatGPT for the full text of a copyrighted book, it refuses and the conversation stops. The same applies to certain image requests. The family's point is that the company built working shut-off switches and pointed them at copyright rather than at a child in crisis.

The complaint also alleges that in his final conversations the product offered to help him write a note to his parents.

Recognize any of this from your own family’s experience? 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.

Request a Case Review →

The rules OpenAI changed

The family amended their complaint in October 2025 to say this was deliberate rather than careless, and that change is what the case now turns on.

Under OpenAI's 2022 rules, ChatGPT had to refuse any discussion of suicide or self-harm outright. It would say it could not help with that, and stop.

On May 8, 2024, five days before the version Adam used was released, that rule was removed. The replacement told the product to give users a space to feel heard instead of ending the conversation. On February 12, 2025, about two months before Adam died, suicide and self-harm were taken off the banned list entirely and moved to a category telling the product to take extra care.

The complaint says Adam's use climbed sharply after that second change, from a few dozen conversations a day in January to more than three hundred a day.

It also says the safety testing behind that release was compressed into about a week so the product could launch ahead of a competitor, and that the tests used single questions rather than long back-and-forth conversations, which is how people actually use it.

OpenAI denies responsibility. Its formal response argued the harm came from using the product in ways its terms do not permit, and that its safety systems and warnings were adequate. That filing drew heavy public criticism.

What the family is asking for

The complaint makes seven legal claims. Behind the labels, they come down to a few arguments.

They are asking for more than money. Matthew and Maria Raine have asked the court to force OpenAI to build in real protections for children and other vulnerable users, which would apply to everyone using the product.

What has changed since

Whatever happens to the case, some things changed after it was filed. In the weeks that followed, OpenAI announced parental controls for ChatGPT, changes to how the product responds in sensitive conversations, and a system that routes some conversations to models with stricter limits.

The case also became a reference point in legislative debates about AI safety rules for children, and it drew the first sustained public attention to who is responsible when an AI product is involved in a death.

Why this case matters to other families

A note for anyone who found this page because of their own child rather than an interest in the case.

Raine is one family's individual lawsuit. It is not a class action, so there is no group to join and nothing to sign up for here. Other families in similar situations have filed their own separate cases, and our OpenAI lawsuits tracker lists what has been filed so far and what each case is about.

Whether any particular situation supports a claim depends on the specific facts, and that is something to talk through with a lawyer rather than decide from an article. A case review through Lawsuit Center is free and carries no obligation.

If you are worried about someone right now, that comes first. The 988 Suicide and Crisis Lifeline is free, confidential, and available at any hour by call or text, and the Crisis Text Line can be reached by texting HOME to 741741. Our Family AI Safety Guide covers what to look for and how to start a conversation with a teenager about chatbot use.

Keep the chat history

There is one practical thing worth acting on, and this case shows why. Adam's parents found the conversations after he died, and those records are what the case is built on. Without them there would be no case and no explanation.

If there is any chance those records matter in your situation, the most important thing is not to delete them.

Closing the account, or letting someone else close it, can take all of it at once. If you are unsure whether any of it will matter, keep it and decide later. What to do in a particular situation is a question for a lawyer rather than an article.

Our guide on what to do in the first days walks through this step by step, including the account, the device and the records worth keeping alongside the conversation.

Case Timeline

Common Questions About Raine v. OpenAI

What is the current status of the case?

The case is still in its early stages in San Francisco County Superior Court. The lawyers are exchanging evidence and arguing preliminary motions. No trial date has been set and no settlement has been announced.

Has there been a verdict?

No. The case has not gone to trial, so there is no verdict, and no court has decided whether OpenAI did anything wrong.

Who are the defendants?

OpenAI, chief executive Sam Altman personally, and unnamed employees and investors. Microsoft is named in other filings as having pressured OpenAI to ship faster, but it is not a defendant here.

Is this a class action?

No. Raine is one family's own lawsuit, brought by Adam's parents and on behalf of his estate. There is no group to join. The other OpenAI cases are separate individual suits that make similar arguments, and the California ones are now managed together for the shared work.

How does Raine relate to the other OpenAI lawsuits?

It came first, and the families who sued later built on it. The Tumbler Ridge school shooting cases, the FSU shooting case, and the Scott overdose case all make versions of the same two arguments: that ChatGPT should be treated as a product, and that OpenAI removed or never applied protections it was capable of. Our OpenAI lawsuits tracker lists them all.

Sources and further reading

Related reading

Attorney Advertising. Lawsuit Informer is operated by a California-licensed attorney. Past results do not guarantee future outcomes.

Affected by harm involving an AI chatbot? If your family experienced serious harm following sustained AI chatbot use, you can request a free, confidential case review through Lawsuit Center. Reviews are conducted by participating legal professionals and intake partners. Submitting a request does not create an attorney-client relationship.

Request a Case Review →

Educational commentary only. Not legal advice. No attorney-client relationship is created.