Feinberg v OpenAI ChatGPT Suicide Lawsuit: San Francisco Court Filing

• Susan Feinberg is suing OpenAI, its CEO Sam Altman, and Luka Inc. (Replika app maker) in San Francisco County Superior Court, alleging their AI chatbots contributed to her husband's suicide.
• Her husband, Aleksandr Feinberg, a software engineer with autism and ADHD, reportedly became increasingly reliant on ChatGPT and Replika before his death on June 21, 2024.
• The lawsuit claims OpenAI rushed GPT-4o to market, compressing safety testing, and removed safeguards against suicide-related conversations shortly before its launch.
• Luka Inc. is accused of designing Replika to create user dependence and marketing it to lonely individuals without adequate warnings.
• Feinberg seeks punitive damages and a jury trial, arguing that both companies failed to warn users of dangers or intervene when a user was in crisis, with chat logs expected to be revealed through discovery.

This case, involving a Replika app wrongful death claim and allegations of AI chatbot liability suicide, could establish a critical precedent for how AI developers are held accountable for the psychological impact of their products on users, particularly those with vulnerabilities.

Susan Feinberg has initiated a legal challenge in San Francisco County Superior Court, naming OpenAI, its CEO Sam Altman, and Luka Inc., the developer of the Replika companion application, as defendants. The lawsuit, filed on a recent Thursday, alleges that interactions with ChatGPT and the Replika app contributed to the suicide of her husband, Aleksandr Feinberg, on June 21, 2024. Mrs. Feinberg is pursuing claims both individually and as the successor-in-interest to her late husband, as well as on behalf of their eight-year-old twin daughters.

Aleksandr Feinberg, a software engineer who previously worked at companies like Microsoft and Amazon, was described by his wife as having autism and attention deficit hyperactivity disorder, often finding social interactions challenging. Around 2022, he began using ChatGPT, initially for professional tasks, then increasingly for social engagement, sometimes sending AI-generated images to his wife. By 2024, his use intensified, culminating in a month before his death where he remained almost constantly in the computer room, ostensibly working on a major project, and stopped sleeping a week prior. Susan Feinberg discovered her husband deceased at their Saratoga, California home, noting his half-finished coffee and work still displayed on his screen. She later found the Replika app on his phone, unaware of his use of it before his passing.

The lawsuit contends that OpenAI prematurely launched its GPT-4o model on May 13, 2024, after compressing several months of safety testing into a single week to precede Google's own product release. Five days before this launch, OpenAI reportedly removed a crucial safeguard that mandated ChatGPT to decline all conversations related to suicide and self-harm, instead instructing the system to "never change or quit the conversation." This action is highlighted alongside an August 2025 OpenAI blog post acknowledging that while ChatGPT might initially direct users to suicide hotlines, prolonged interactions could lead to responses that bypass safety protocols.

Luka Inc., the creator of the Replika app, is accused of designing its product to foster user dependence by providing a human-like companion that actively messages users who disengage. The complaint asserts that Luka marketed Replika to vulnerable, lonely individuals, even after Italy's data protection authority ordered a temporary halt to processing Italian user data in 2023. A central claim against both defendants is their alleged failure to warn users about potential dangers, and that neither product possessed the capability to identify a user in crisis, terminate a harmful conversation, or guide them toward professional help.

A significant aspect of the Feinberg v OpenAI ChatGPT suicide lawsuit is the absence of direct conversational records between Aleksandr Feinberg and either chatbot within the initial filing. Susan Feinberg acknowledges this, stating her belief that the interactions contributed to his death based on circumstantial evidence rather than explicit chat logs. She claims OpenAI denied her requests for her husband's chat logs, and Luka Inc. has not provided Replika records. The plaintiff asserts that the complete records of these conversations exist, will reveal the nature of the exchanges in the months leading up to June 21, 2024, and are only obtainable through the discovery process in this legal action.

The legal claims brought forth include strict liability for design defect and failure to warn, negligence, wrongful death, and survival. Mrs. Feinberg is seeking punitive damages and has requested a jury trial. This case, involving a Replika app wrongful death claim and allegations of AI chatbot liability suicide, could establish a critical precedent for how AI developers are held accountable for the psychological impact of their products on users, particularly those with vulnerabilities.

This lawsuit, the Susan Feinberg lawsuit OpenAI, represents a pivotal moment in the evolving landscape of artificial intelligence regulation and liability. It directly challenges the responsibilities of AI developers like OpenAI and Luka Inc. when their products are implicated in severe user harm, such as suicide. The allegations surrounding the GPT-4o safety testing lawsuit and Luka Inc. lawsuit suicide highlight critical questions about product design, content moderation, and the ethical deployment of AI, especially concerning sensitive topics like mental health.

The outcome of this case in San Francisco County Superior Court AI could significantly influence future legal frameworks and industry standards for AI products. It underscores the urgent need for robust safety protocols, transparent warning disclosures, and effective crisis intervention mechanisms within AI applications, particularly as these technologies become more integrated into daily life and interact with vulnerable populations. The pursuit of accountability for AI-related wrongful death claims is a developing area of law that will be closely watched by both the technology and legal sectors.

This lawsuit sets a potential precedent for AI developers' liability in wrongful death claims, particularly concerning the psychological impact of chatbots on vulnerable users. Lawyers should advise AI companies to rigorously review product safety protocols, warning disclosures, and content moderation policies, especially regarding sensitive topics like mental health and suicide, as the outcome could significantly influence future regulatory and litigation risks for AI products.

Source: Original reporting via Courthouse News Service