News & Analysis

OpenAI School Shooting Lawsuits: What the Families Claim

By David Meldofsky

Published May 2, 2026 · Updated September 2, 2026

OpenAI is being sued over two mass shootings. Families and survivors say the company knew ChatGPT was being used to plan violence and did not stop it. This page explains who is suing, what they claim, and whether any of it applies to you.

What happened

Two shootings are involved, in two different countries.

On April 17, 2025, a gunman killed two people at Florida State University. On February 10, 2026, an 18-year-old killed eight people in Tumbler Ridge, British Columbia. Five of them were children at the local school, along with an education assistant. Two more students were severely wounded, one of them left with life-changing brain injuries. The attacker died by suicide.

In both cases, the person who carried out the attack had been using ChatGPT beforehand.

Who is suing OpenAI

Two separate groups, in two separate courts.

The Tumbler Ridge families and survivors. Seven families of the dead and injured sued in San Francisco federal court in April 2026. A second group started filing in September 2026, made up of students, teachers, and a principal who were inside the school during the attack. Their lawyers have said that group will grow to about thirty. Our Tumbler Ridge case guide covers those filings.

Families and survivors from FSU. Three cases have been filed in Florida federal court. Tiru Chabba's family sued in May 2026. Reese Gourley, a student who was shot and survived, sued in August. Betty Morales sued the same month over the death of her husband Roberto, the other person killed. A fourth case, by another student who was shot, is in Florida state court. Our FSU case guide covers them.

Florida's attorney general is also involved separately. He opened a criminal investigation and filed the state's own lawsuit against OpenAI in June 2026. An investigation is a set of questions rather than a conclusion, and no criminal charges have been filed.

What the lawsuits say OpenAI did wrong

The claims differ between the two case sets, and the difference matters.

The Tumbler Ridge lawsuits say OpenAI knew about this specific person ahead of time. According to the complaints, an automated system flagged his account in June 2025, eight months before the attack, because of what he was writing about guns. Staff reviewed it, decided he was a real danger, and recommended calling the Royal Canadian Mounted Police. The complaints say company leadership overruled them, shut the account down, and never made the call. The person then opened a new account and kept using ChatGPT.

In April 2026, Sam Altman published a letter to the Tumbler Ridge community saying he was deeply sorry the company did not alert police about the account. OpenAI disputes the rest, and says the description of who made that decision, and why, is false.

The FSU lawsuit does not claim OpenAI knew about that shooter in advance. It argues the product itself was built carelessly, pointing to chat records police recovered afterward and to reporting that safety testing was rushed.

Does this affect me?

The lawsuits on this page were brought by people who were at the two schools and by families of the people killed there. That is who these particular cases are about.

They are not the only cases against OpenAI. A separate and larger group was brought by people who say ChatGPT harmed them or someone in their family directly, involving suicides, overdoses, and mental health crises following long conversations with the product. If that is closer to your situation, our OpenAI lawsuits tracker lists every case and what each one is about.

Whether any particular situation supports a claim turns on the specific facts, and that is what a case review is for. It is free and carries no obligation.

If someone in your life is in immediate danger, contact emergency services rather than a law firm. The resources at the top of this page are free and available around the clock.

Keep the chat history

One practical note, because it has a deadline. These cases run on the conversations themselves. In Raine the family found the chat history after their son died and it became the center of the complaint, and in the FSU case the records came from police.

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

• Deleting is permanent. OpenAI says a deleted conversation leaves the account straight away and is removed from its systems within 30 days, and that it cannot be recovered afterward through the app, the support team, or anywhere else.
• Archiving is not deleting. Archived conversations stay in the account.
• You can request a copy. An account holder can ask OpenAI for an export of the account's data through settings or its privacy portal. It takes up to seven days, and it will not bring back anything already deleted.
• Memory is stored separately. If the memory feature was switched on, ChatGPT saved details about the person across conversations. Those saved memories sit apart from the chat history, so deleting conversations does not clear them.

Closing the account can take all of it at once. If you are unsure whether any of it will matter, keep it and decide later.

Why these cases are hard to win

Three problems stand in the way, and they apply to both case sets.

Proving the product made a difference. OpenAI will argue that these attacks had many causes and would have happened regardless of ChatGPT. So it is not enough for the families to show the product was used. They have to show it changed what happened. No one can run the events again to find out, which means this turns on what the chat records contain and what experts make of them. It is the hardest part of both cases, and it is where they can still fail even if every other question goes the families' way.

Deciding what ChatGPT legally is. Courts have not settled whether an AI's answers count as a product, like a defective car part, or as something closer to speech. The answer changes which rules apply and how much protection OpenAI gets. Our explainer on whether AI output is a product or content covers this in plain terms.

Whether OpenAI owed these people anything. A company is not usually responsible for what a stranger does to someone else. Each case set has a different way around that, and the difference is worth understanding.

In Tumbler Ridge, the argument is that OpenAI's own staff flagged the account and said to call police, so the company was warned about this specific person and did nothing.

In the FSU case, no one at OpenAI is said to have flagged the shooter beforehand. The argument there is that the warning signs were sitting in his chat history all along, that the company should have caught them, and that letting him keep using the product was itself the failure. That claim has a name, negligent entrustment, and it usually applies to handing someone a car or a gun. Whether it stretches to a software account is an open question, and it is one of the more closely watched parts of that case.

What happens next, and when

The next step in both case sets is a motion to dismiss. That is OpenAI asking the judge to throw the cases out before any evidence is exchanged, and it is standard in cases like these.

If the cases survive that, the lawyers get access to OpenAI's internal records. In the Tumbler Ridge cases they have said they want the chat logs, along with the internal documents showing who recommended calling police and who decided not to.

These early rulings matter well beyond these families. They will shape what every AI company can be held responsible for, which is why the motions, rather than any eventual verdict, are the thing to watch. Expect this to take years.

Common questions

What are the OpenAI school shooting lawsuits?

Two sets of lawsuits against OpenAI over mass shootings. One set comes from families and survivors of the February 2026 school shooting in Tumbler Ridge, British Columbia, filed in California federal court starting in April 2026. The other was filed in Florida in May 2026 by the family of someone killed in the April 2025 Florida State University shooting. Nothing has been proven and OpenAI denies wrongdoing.

How are the Tumbler Ridge and FSU cases different?

The Tumbler Ridge cases say OpenAI knew about that specific person before the attack, because its own staff flagged his account and recommended calling police. The FSU case makes no such claim. It argues instead that the product was built carelessly, based on chat records recovered after the shooting.

What will be hardest for the families to prove?

Causation. Showing that ChatGPT actually contributed to a mass shooting, rather than being one factor among many, is harder than any of the legal arguments in these cases. Even if the families win the legal questions, this is where the cases can still fail.

Can I sue OpenAI if I used ChatGPT?

Whether someone has a claim depends on their own facts, and no article can answer that. What can be said is that people have sued. Some of the cases on this docket were brought by people who used ChatGPT and by their families, not only by people caught up in these two shootings. Our tracker lists what has been filed so far. If you think your situation resembles any of them, a case review is free and carries no obligation.

Sources and further reading

• Rushlow v. Altman complaint (PDF, 48 pages), the first available second-wave Tumbler Ridge filing
• Joshi v. OpenAI Foundation complaint (PDF), the first FSU case
• Associated Press: coverage of the Tumbler Ridge lawsuits
• NPR: survivors of the Tumbler Ridge shooting sue OpenAI
• The Globe and Mail: Tumbler Ridge survivors sue OpenAI
• OpenAI Help Center: deleting and archiving chats
• OpenAI Help Center: exporting your data

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