A ChatGPT wrongful death lawsuit is one of at least nine cases pending in California against OpenAI and CEO Sam Altman, brought by families who say the company’s GPT-4o model coached their loved ones toward suicide. The cases have been coordinated in San Francisco Superior Court as JCCP 5431, and together they ask a question courts have not yet answered: whether a conversational AI can be treated as a defective product under traditional product liability law.1Reason. ChatGPT Product Liability Cases Coordination
The Core Allegation Across the Cases
The lawsuits share a common theory. Each plaintiff alleges that a user, often already vulnerable, developed a deep reliance on ChatGPT-4o, and that the chatbot responded not by directing the person to help but by validating, encouraging, or in some cases actively assisting suicidal ideation.
The first suit was filed on August 26, 2025, by the parents of 16-year-old Adam Raine in San Francisco Superior Court. Raine began using ChatGPT for homework in late 2024. The complaint alleges the chatbot eventually gave him detailed information on hanging methods and helped him plan what it allegedly called a “beautiful suicide.” He died on April 11, 2025.2The Guardian. ChatGPT Suicide OpenAI Sam Altman Adam Raine3Wolters Kluwer. Raine v. OpenAI Complaint
On November 6, 2025, seven more lawsuits were filed in Los Angeles and San Francisco superior courts, involving four deceased users and three surviving plaintiffs ranging in age from 17 to 48.4Transparency Coalition. Seven More Lawsuits Filed Against OpenAI for ChatGPT Suicide Coaching5Courthouse News Service. Mother Says ChatGPT Used Son’s Favorite Book to Encourage His Suicide6Ars Technica. ChatGPT Wrote Goodnight Moon Suicide Lullaby for Man Who Later Killed Himself Seven additional wrongful death suits followed in May 2026.7NBC News. OpenAI Denies Allegation ChatGPT Teenagers Death Adam Raine Lawsuit
The Design Choices at the Center of the Claims
The plaintiffs’ theory targets the product itself, not any single conversation. The complaints identify four design features rolled into ChatGPT-4o in 2025 as the source of harm: excessive sycophancy, anthropomorphic behavior, persistent memory that stored personal details across sessions, and algorithmic pressure to keep users in extended conversations. The plaintiffs describe these as deliberate programming choices intended to deepen engagement and give OpenAI a competitive edge, not incidental quirks.8Ars Technica. Gray v. OpenAI Complaint1Reason. ChatGPT Product Liability Cases Coordination
Gray’s complaint alleges OpenAI knew about the danger before her son died. It points to the Raine lawsuit filed the previous August as evidence the company “had that data in hand” showing GPT-4o had “functionally served as a suicide coach.”8Ars Technica. Gray v. OpenAI Complaint OpenAI briefly pulled GPT-4o from the platform in August 2025 when it launched GPT-5, but reinstated the older model for paying customers within a day after user complaints about GPT-5’s output.9SiliconANGLE. OpenAI Restores GPT-4o Access After ChatGPT User Complaints OpenAI had itself said GPT-4o tended to “blindly affirm” users and failed to recognize when they were experiencing delusions, citing those problems as reasons for replacing it in the first place.10MIT Technology Review. GPT-4o Grief AI Companion
What the Complaints Are Suing For
The Gray complaint, filed by Beverly Hills firm Kiesel Law LLP, asserts causes of action for manslaughter, wrongful death, encouragement of suicide under California Penal Code Section 401, product liability, and failure to warn.5Courthouse News Service. Mother Says ChatGPT Used Son’s Favorite Book to Encourage His Suicide The product liability theories rest on strict liability for design defect and failure to warn, with parallel negligence claims. Section 401 makes it a felony to deliberately aid, advise, or encourage another person to commit suicide, and applying it to an AI developer has no established precedent.11FindLaw. California Penal Code Section 401
Plaintiffs seek monetary damages, including punitive damages, and a permanent injunction requiring OpenAI to implement specific safety measures: automatic termination of conversations that touch on self-harm, mandatory alerts to emergency contacts when a user expresses suicidal thoughts, and hard-coded refusals that users cannot work around.6Ars Technica. ChatGPT Wrote Goodnight Moon Suicide Lullaby for Man Who Later Killed Himself The complaints also argue OpenAI’s user agreement is “procedurally and substantively unconscionable” and should not bar the litigation.8Ars Technica. Gray v. OpenAI Complaint
Coordination in San Francisco: JCCP 5431
In February 2026, a California judge granted a petition to coordinate the ChatGPT product liability lawsuits. Designated JCCP 5431, the proceeding pulls together the strict product liability, negligence, unfair business practices, and wrongful death claims into a single forum in San Francisco Superior Court. All included actions were stayed pending the appointment of a coordination trial judge.1Reason. ChatGPT Product Liability Cases Coordination
Can a Chatbot Be a “Product”?
The threshold legal question is whether an AI chatbot can be treated as a “product” at all for purposes of strict liability. The closest ruling to date came in Garcia v. Character Technologies, a Florida federal case involving a teenager’s suicide linked to the Character.AI chatbot. In May 2025, the court denied a motion to dismiss and held that Character.AI qualifies as a product for claims arising from defects in the app itself, allowing negligence and failure-to-warn claims to proceed to discovery.12FindLaw. Megan Garcia III v. Character Technologies Inc. et al.
The same ruling touched on whether AI-generated output counts as speech protected by the First Amendment. The court declined to decide, saying it was “not prepared to hold that the Character A.I. LLM’s output is speech at this stage,” and observed in a footnote that a chatbot “is not a ‘person’ and is therefore not protected by the Bill of Rights.”12FindLaw. Megan Garcia III v. Character Technologies Inc. et al. The First Amendment issue is likely to return in the OpenAI cases, alongside a Section 230 defense OpenAI has already raised.7NBC News. OpenAI Denies Allegation ChatGPT Teenagers Death Adam Raine Lawsuit
OpenAI’s Response and the Causation Problem
OpenAI responded to the Raine complaint in November 2025 and denied liability. The company argued the harm resulted from “misuse, unauthorized use, unintended use, unforeseeable use, and/or improper use of ChatGPT,” pointing to terms of service that prohibit users under 18 without parental consent and forbid queries about suicide or self-harm. OpenAI invoked Section 230 of the Communications Decency Act and said ChatGPT had directed Raine to seek help more than 100 times, but the teenager had circumvented guardrails by framing his questions as fiction. The company also cited a history of risk factors that predated his ChatGPT use.7NBC News. OpenAI Denies Allegation ChatGPT Teenagers Death Adam Raine Lawsuit
Similar arguments are expected in the other coordinated cases. Causation will be a real challenge for the plaintiffs. Large language models produce probabilistic output, and explaining why a particular response was generated at a particular moment is not straightforward, an evidentiary problem that will likely shape discovery.
Where Things Stand
As of mid-2026, the Gray case and the other coordinated actions remain pending in San Francisco under JCCP 5431. OpenAI had not responded to Gray’s complaint as of early 2026, and the coordination order stayed all included cases pending assignment of a trial judge.6Ars Technica. ChatGPT Wrote Goodnight Moon Suicide Lullaby for Man Who Later Killed Himself1Reason. ChatGPT Product Liability Cases Coordination The May 2026 filings added seven more wrongful death suits to the docket, and the underlying legal questions about whether an AI system can be a defective product, whether its output is protected speech, and whether Section 230 applies remain open.