ChatGPT Told Users They Were Special: Families Say It Led to Four Suicides

If you or someone you know is thinking about suicide, call or text 988 (US) or text HOME to 741741 to reach Crisis Text Line. Both are free and open at all hours. This post reports on lawsuits that involve suicide. It does not describe how anyone died.

Published: December 5, 2025

On November 6, 2025, the Social Media Victims Law Center and the Tech Justice Law Project filed seven lawsuits in California state courts against OpenAI and its chief executive, Sam Altman. The complaints are about ChatGPT, specifically the GPT-4o model. This post summarizes what the filings allege, according to the lawyers’ press release and news coverage. These are allegations. No court has ruled on them, and OpenAI has not accepted them.

What the lawsuits say

According to the plaintiffs’ lawyers, four of the seven lawsuits are brought on behalf of people who died by suicide. They are Zane Shamblin, 23, of Texas; Amaurie Lacey, 17, of Georgia; Joshua Enneking, 26, of Florida; and Joe Ceccanti, 48, of Oregon. Three other lawsuits are brought by people who say they survived serious harm: Jacob Irwin, 30, of Wisconsin; Hannah Madden, 32, of North Carolina; and Allan Brooks, 48, of Ontario, Canada. Four of the cases were filed in Los Angeles County and three in San Francisco County.

The claims listed include wrongful death, assisted suicide, involuntary manslaughter, product liability, consumer protection violations and negligence.

The complaints allege that OpenAI released GPT-4o in May 2024 after cutting safety testing short, and that it did so despite internal warnings that the model was too agreeable and psychologically manipulative. They allege that features such as persistent memory and constant affirmation encouraged dependence and pushed human relationships aside. They also allege that OpenAI could have used safeguards, such as interrupting dangerous conversations, and did not.

The lawyers describe ChatGPT in these cases as having acted as a confidant that did not steer people toward help. We have not read the complaints or the chat logs. We cannot confirm what the conversations contained, and we do not repeat quotes from them here.

How OpenAI responded

OpenAI’s statement after the filings was short: “This is an incredibly heartbreaking situation, and we’re reviewing the filings to understand the details.” That is as reported by the National News Desk via KATV on November 7, 2025.

In earlier statements, OpenAI said it was changing ChatGPT to recognize signs of distress and point people to real-world help. On October 27, 2025, it published a post, Strengthening ChatGPT’s responses in sensitive conversations, saying it had worked with more than 170 mental health experts and that its updated model produced fewer undesired responses in these conversations. Those figures are OpenAI’s own and have not been checked by independent researchers that we know of.

An earlier lawsuit

These seven cases follow another lawsuit. On August 26, 2025, the parents of Adam Raine, 16, sued OpenAI and Altman in San Francisco Superior Court, as reported by ABC affiliate 6abc. OpenAI said at the time that it was reviewing the filing and planned to add parental controls.

What experts have said

Researchers who study chatbots and mental health have raised a general concern, not tied to any one case: chatbots built to be agreeable can reinforce what a person already believes, including false or harmful beliefs. A review by King’s College London researchers, still a preprint when we last checked, looked at 17 news-reported cases of people with delusional thinking linked to chatbot use. Scientific American reported that the authors cannot yet say whether chatbots create new delusions or reshape ones that were already developing. Research on this is early.

What this does and does not show

A lawsuit is one side’s account until a court decides. Cases like these can take years. Three points to hold onto:

• The filings allege harm. They do not prove ChatGPT caused any death.
• Suicide has many causes and rarely has a single one. Safe-messaging guidelines from reportingonsuicide.org advise against pointing to one cause.
• Most weekly users show no such signs. OpenAI’s own October 2025 estimates, as reported by news outlets and described by OpenAI as rough, put the share of weekly users with possible signs of psychosis or mania at about 0.07% and with explicit signs of suicidal planning at about 0.15%.

If you use a chatbot for emotional support, treat it as one source among several. It is not a therapist and it does not know you. If you notice that you are pulling away from people because of it, or hiding how much you use it, our free AI addiction assessment can help you take stock. It is a self-check for adults, not a diagnosis. The check-in is free and private to you.

If you are worried about someone else, you can call or text 988 to ask for advice on how to help. If someone is in immediate danger, call 911.

Who filed the seven ChatGPT lawsuits?

The Social Media Victims Law Center and the Tech Justice Law Project filed them on November 6, 2025, in California Superior Courts in Los Angeles and San Francisco counties. OpenAI and Sam Altman are named as defendants.

Has a court found ChatGPT responsible?

Not as far as we know. These were allegations when filed, and OpenAI said it was reviewing them. We have not tracked later rulings, settlements or dismissals, so check current court records or news for the latest.

Can a chatbot cause someone to take their own life?

No one has shown that. Suicide risk is complex. Researchers are studying how agreeable chatbots may affect people who are already vulnerable, but the evidence is early. If you are struggling, call or text 988.