Families of people who died by suicide or other causes, and survivors of self-harm, are covered by roughly two dozen individual lawsuits alleging OpenAI’s ChatGPT contributed to those deaths and injuries. No class has been certified, there is no settlement, and there is nothing to file or claim.
Allegations Only · No Settlement
This article describes lawsuits. The statements about ChatGPT and OpenAI are unproven allegations made by the plaintiffs. OpenAI denies them, no court has found OpenAI liable, and there is no class and nothing to claim. This page is informational and is not legal advice.
What Is This About?
Since August 2025, families and individuals have filed a series of lawsuits alleging that ChatGPT, OpenAI’s chatbot, deepened users’ suicidal thinking or delusions instead of steering them toward help, and that the design of the GPT-4o model made those failures more likely. The suits name OpenAI and, in most cases, its chief executive, Sam Altman. Some also name Microsoft.These are individual cases, not a class action. Most of the ones filed in California state courts have been coordinated before a single judge in San Francisco Superior Court as Judicial Council Coordination Proceeding No. 5431, In re ChatGPT Product Liability Cases. Out of respect for the families, this page describes each case by court, filing date and the person’s age rather than by name.
Free settlement alerts
Get notified when new class actions open to claims
Join thousands of readers who get the latest class action settlements you may qualify for — delivered straight to your inbox.
What the Lawsuits Allege
The complaints share a core theory. They allege that OpenAI released GPT-4o in May 2024 with design choices meant to keep people engaged, including a habit of agreeing with and validating users, and that it loosened the model’s instructions on self-harm conversations before launch. According to the plaintiffs, ChatGPT continued long conversations with users who expressed suicidal thoughts or delusional beliefs, discouraged some from confiding in family, and did not reliably end those conversations or direct them to help.The legal claims vary by case but commonly include wrongful death, strict product liability for design defect and failure to warn, negligence, and violations of California’s Unfair Competition Law. Several of the November 2025 complaints also pleaded assisted suicide and involuntary manslaughter. Some plaintiffs ask for court orders requiring OpenAI to change ChatGPT, such as automatically ending conversations about self-harm, in addition to damages.
The Cases So Far
Based on court filings, the plaintiffs’ lawyers’ announcements and news coverage, the publicly reported cases include:- August 2025, San Francisco Superior Court: the parents of a 16-year-old California boy who died in April 2025 filed the first wrongful-death suit against OpenAI and Altman. They amended it in October 2025 to allege intentional misconduct.
- November 2025, seven suits in Los Angeles and San Francisco Superior Courts: filed together by the Social Media Victims Law Center and Tech Justice Law Project. Four involve deaths, of a 17-year-old from Georgia and adults aged 23 (Texas), 26 (Florida) and 48 (Oregon). Three involve survivors aged 30 (Wisconsin), 32 (North Carolina) and 48 (Ontario, Canada) who allege psychological harm.
- December 2025, San Francisco Superior Court and federal court in San Francisco: suits over a Connecticut murder-suicide allege ChatGPT reinforced a man’s paranoid delusions before he killed his mother and himself. In April 2026, a federal judge refused OpenAI’s request to dismiss or pause the federal case in favor of the state one.
- January 2026, Los Angeles Superior Court: the family of a 40-year-old Colorado man.
- May 2026, San Francisco Superior Court: the family of a 19-year-old who died of an accidental overdose, alleging ChatGPT gave dangerous drug advice. See OCA’s coverage of that case.
- June 2026, San Francisco Superior Court: the mother of a 24-year-old Canadian woman who died in July 2025, and, days later, the family of a 29-year-old Alabama woman.
- July 2026, San Francisco Superior Court: a 34-year-old California man who survived, seeking damages and an order requiring safety changes to ChatGPT.
The list is not complete. By late August 2026, roughly two dozen cases had been gathered into the coordinated proceeding, about half of them wrongful-death claims, and the plaintiffs’ lawyers have said more are being prepared.
Where the Coordinated Cases Stand
The San Francisco court coordinated the first group of cases in early 2026. On August 4, 2026, Judge Ethan P. Schulman entered his first case management order appointing four co-lead lawyers, a liaison counsel and a steering committee for the plaintiffs. According to reporting on the proceeding, the judge allowed evidence-gathering to begin in late September 2026, and OpenAI’s initial responses to the complaints were expected by October 30, 2026.No trial dates have been set, and no ruling on OpenAI’s legal defenses had been reported as of October 10, 2026. Coordinated proceedings of this kind typically move toward a small number of test, or "bellwether," trials, but none has been scheduled.
How OpenAI Has Responded
OpenAI denies that ChatGPT caused these deaths and injuries. In its November 2025 answer to the first case, it argued that the user had misused ChatGPT and violated its terms of use, that his death was not caused by the product, that ChatGPT had directed him to seek help more than 100 times, and that Section 230 of the Communications Decency Act protects it. It filed the full chat history under seal. OpenAI has also argued that ChatGPT is a service rather than a product, which would affect whether product-liability law applies.Publicly, the company has called the cases heartbreaking and pointed to a series of changes:
- September 2025: parental controls that let parents link a teen’s account, set quiet hours and receive alerts in rare cases of acute distress.
- October 2025: updated model guidelines on mental health and emotional reliance.
- January 2026: an age-prediction system on consumer plans that applies teen protections to users it estimates are under 18.
- February–April 2026: the retirement of GPT-4o from ChatGPT.
- 2026: a "trusted contact" feature for adults, announced in February.
Related Cases
Separate lawsuits filed in 2026 allege that ChatGPT helped the gunman in the April 2025 shooting at Florida State University plan his attack. At least four are now in the U.S. District Court for the Northern District of Florida in Tallahassee: a widow’s wrongful-death suit filed in May 2026 (No. 4:26-cv-00222), two suits filed in late August 2026, and an injured student’s suit that began in Leon County circuit court and was moved to federal court on September 17, 2026. Florida’s attorney general sued OpenAI in June 2026 on separate consumer-protection and product-liability grounds; see OCA’s explainer on the Florida lawsuit. OpenAI disputes those claims as well.A larger group of federal suits in San Francisco arises from the February 2026 school shooting in Tumbler Ridge, British Columbia. More than 30 victims’ families and survivors allege OpenAI flagged the shooter’s ChatGPT conversations months earlier but did not alert police, and the Province of British Columbia sued OpenAI and Altman on September 21, 2026 (No. 3:26-cv-10743, N.D. Cal.). OpenAI has asked the court to send at least some of those cases to Canada.
Lawsuits against Character.AI and Google over teen chatbot use, including a Florida wrongful-death case, settled on confidential terms in January 2026. That history is covered in Character.AI & Google Settle Teen Chatbot Harm Lawsuits.
Questions
Is there a ChatGPT class action for suicide or self-harm?
No. These are individual wrongful-death and personal-injury lawsuits, most of them coordinated for pretrial purposes in San Francisco Superior Court as JCCP No. 5431. A coordinated proceeding is not a class action: there is no class, no settlement fund and nothing for the public to join or claim. Each family or injured person brings their own case.
What is JCCP 5431?
Judicial Council Coordination Proceeding No. 5431, In re ChatGPT Product Liability Cases, groups ChatGPT injury and wrongful-death cases filed in California state courts before one judge, Judge Ethan P. Schulman of San Francisco Superior Court, for shared pretrial work such as discovery and motions. It was created in early 2026 and had grown to roughly two dozen cases by August 2026.
How has OpenAI responded to the lawsuits?
OpenAI denies that ChatGPT caused the deaths and injuries alleged. In its November 2025 answer to the first case, it raised defenses including misuse of the product, violations of its terms of use, lack of causation, comparative fault and Section 230 of the Communications Decency Act, and it has argued that ChatGPT is a service rather than a product. It has also called the cases heartbreaking and pointed to safety changes it has made.
Has any ChatGPT wrongful-death case gone to trial or settled?
Not as of October 10, 2026. No trial dates had been set in the coordinated cases, no ruling on OpenAI’s legal defenses had been reported, and no settlement had been announced. The separate Character.AI and Google teen cases did settle on confidential terms in January 2026.
Sources
- Tech Justice Law Project — ChatGPT product liability cases coordinated (JCCP 5431)
- MLex — Case Management Order No. 1, August 4, 2026
- Lawyer Monthly — status of the coordinated cases, October 2026
- NBC — OpenAI’s November 2025 answer in the first case
- Tech Justice Law Project — the seven November 2025 lawsuits
- CBS News — Connecticut murder-suicide lawsuit
- Courthouse News — federal court keeps the Connecticut case, April 2026
- Al Jazeera — June 2026 suit over a Canadian woman’s death
- Tech Justice Law Project — July 2026 survivor lawsuit
- Fox 49 — FSU shooting lawsuits
- CourtListener — No. 4:26-cv-00222 (N.D. Fla.)
- Al Jazeera — British Columbia sues OpenAI over the Tumbler Ridge shooting
- CourtListener — No. 3:26-cv-10743 (N.D. Cal.)
- OpenAI — "Our approach to mental health-related litigation," November 2025
- OpenAI — parental controls, September 2025
- OpenAI — retiring GPT-4o
